BCU_Classifier.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "G8ZYoZzs-gkR"
      },
      "outputs": [],
      "source": [
        "from astropy.table import Table\n",
        "from scipy.stats import zscore\n",
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "from sklearn.linear_model import LinearRegression\n",
        "import sys\n",
        "import os\n",
        "from scipy.stats import gaussian_kde\n",
        "from livelossplot import PlotLossesKeras\n",
        "import numpy as np\n",
        "import json\n",
        "import urllib\n",
        "import re\n",
        "sns.set(style=\"darkgrid\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Random seed set as 0\n"
          ]
        }
      ],
      "source": [
        "from utils import *\n",
        "set_seed(0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "V4jXX_CU-vjZ"
      },
      "outputs": [],
      "source": [
        "df_h = Table.read('/content/drive/MyDrive/BCU Classification/table-4LAC-DR3-h.fits')\n",
        "df_h = df_h.to_pandas()\n",
        "df_h = df_h[df_h['Flags']==0]\n",
        "\n",
        "df_l = Table.read('/content/drive/MyDrive/BCU Classification/table-4LAC-DR3-l.fits')\n",
        "df_l = df_l.to_pandas()\n",
        "df_l = df_l[df_l['Flags']==0]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 342
        },
        "id": "_i_LPjOp-_Od",
        "outputId": "1cb14465-f20a-41ed-e06b-b80821173e59"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-94938362-03e2-47ad-95c8-dd67fbc67bf3\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Source_Name</th>\n",
              "      <th>DataRelease</th>\n",
              "      <th>RAJ2000</th>\n",
              "      <th>DEJ2000</th>\n",
              "      <th>GLON</th>\n",
              "      <th>GLAT</th>\n",
              "      <th>Signif_Avg</th>\n",
              "      <th>Flux1000</th>\n",
              "      <th>Unc_Flux1000</th>\n",
              "      <th>Energy_Flux100</th>\n",
              "      <th>...</th>\n",
              "      <th>HE_EPeak</th>\n",
              "      <th>Unc_HE_EPeak</th>\n",
              "      <th>HE_nuFnuPeak</th>\n",
              "      <th>Unc_HE_nuFnuPeak</th>\n",
              "      <th>nu_syn</th>\n",
              "      <th>nuFnu_syn</th>\n",
              "      <th>Variability_Index</th>\n",
              "      <th>Frac_Variability</th>\n",
              "      <th>Unc_Frac_Variability</th>\n",
              "      <th>Highest_energy</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>b'4FGL J0001.2+4741'</td>\n",
              "      <td>1</td>\n",
              "      <td>0.3126</td>\n",
              "      <td>47.685902</td>\n",
              "      <td>114.250198</td>\n",
              "      <td>-14.338059</td>\n",
              "      <td>4.692600</td>\n",
              "      <td>1.259796e-10</td>\n",
              "      <td>2.877197e-11</td>\n",
              "      <td>1.499454e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>-inf</td>\n",
              "      <td>-inf</td>\n",
              "      <td>1.000000e+14</td>\n",
              "      <td>3.716456e-13</td>\n",
              "      <td>25.313953</td>\n",
              "      <td>0.675882</td>\n",
              "      <td>0.312850</td>\n",
              "      <td>65.575203</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>b'4FGL J0001.2-0747'</td>\n",
              "      <td>1</td>\n",
              "      <td>0.3151</td>\n",
              "      <td>-7.797100</td>\n",
              "      <td>89.032722</td>\n",
              "      <td>-67.305008</td>\n",
              "      <td>23.782486</td>\n",
              "      <td>7.471219e-10</td>\n",
              "      <td>5.041939e-11</td>\n",
              "      <td>8.339171e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>796.297852</td>\n",
              "      <td>766.797424</td>\n",
              "      <td>8.686245e-07</td>\n",
              "      <td>7.492923e-08</td>\n",
              "      <td>9.120108e+13</td>\n",
              "      <td>1.936770e-12</td>\n",
              "      <td>46.780693</td>\n",
              "      <td>0.406565</td>\n",
              "      <td>0.111870</td>\n",
              "      <td>86.975197</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>b'4FGL J0001.4-0010'</td>\n",
              "      <td>3</td>\n",
              "      <td>0.3717</td>\n",
              "      <td>-0.169900</td>\n",
              "      <td>96.891991</td>\n",
              "      <td>-60.491329</td>\n",
              "      <td>4.900018</td>\n",
              "      <td>1.082246e-10</td>\n",
              "      <td>2.914440e-11</td>\n",
              "      <td>1.231385e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>15475.512695</td>\n",
              "      <td>18471.714844</td>\n",
              "      <td>1.494930e-07</td>\n",
              "      <td>8.294155e-08</td>\n",
              "      <td>3.609939e+12</td>\n",
              "      <td>2.473256e-13</td>\n",
              "      <td>9.272764</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>10.000000</td>\n",
              "      <td>30.056200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>b'4FGL J0001.5+2113'</td>\n",
              "      <td>1</td>\n",
              "      <td>0.3815</td>\n",
              "      <td>21.218300</td>\n",
              "      <td>107.649353</td>\n",
              "      <td>-40.167652</td>\n",
              "      <td>54.653084</td>\n",
              "      <td>1.347354e-09</td>\n",
              "      <td>6.172809e-11</td>\n",
              "      <td>2.555889e-11</td>\n",
              "      <td>...</td>\n",
              "      <td>65.680008</td>\n",
              "      <td>18.915735</td>\n",
              "      <td>8.844835e-06</td>\n",
              "      <td>8.229393e-07</td>\n",
              "      <td>1.584893e+14</td>\n",
              "      <td>1.152471e-12</td>\n",
              "      <td>1910.935791</td>\n",
              "      <td>0.996138</td>\n",
              "      <td>0.213478</td>\n",
              "      <td>-inf</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>b'4FGL J0001.8-2153'</td>\n",
              "      <td>3</td>\n",
              "      <td>0.4647</td>\n",
              "      <td>-21.886499</td>\n",
              "      <td>55.184696</td>\n",
              "      <td>-77.572540</td>\n",
              "      <td>4.578068</td>\n",
              "      <td>9.047530e-11</td>\n",
              "      <td>2.442527e-11</td>\n",
              "      <td>1.061114e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>6281.065430</td>\n",
              "      <td>2626.445068</td>\n",
              "      <td>1.545364e-07</td>\n",
              "      <td>5.278574e-08</td>\n",
              "      <td>1.659588e+13</td>\n",
              "      <td>9.716011e-13</td>\n",
              "      <td>24.557972</td>\n",
              "      <td>0.902851</td>\n",
              "      <td>0.380107</td>\n",
              "      <td>-inf</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 41 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-94938362-03e2-47ad-95c8-dd67fbc67bf3')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-94938362-03e2-47ad-95c8-dd67fbc67bf3 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-94938362-03e2-47ad-95c8-dd67fbc67bf3');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-18a80459-0240-40aa-ad2b-b47f90bd9e83\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-18a80459-0240-40aa-ad2b-b47f90bd9e83')\"\n",
              "            title=\"Suggest charts.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "    background-color: #E8F0FE;\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: #1967D2;\n",
              "    height: 32px;\n",
              "    padding: 0 0 0 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: #E2EBFA;\n",
              "    box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: #174EA6;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "    background-color: #3B4455;\n",
              "    fill: #D2E3FC;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart:hover {\n",
              "    background-color: #434B5C;\n",
              "    box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "    filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "    fill: #FFFFFF;\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const charts = await google.colab.kernel.invokeFunction(\n",
              "          'suggestCharts', [key], {});\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-18a80459-0240-40aa-ad2b-b47f90bd9e83 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "            Source_Name  DataRelease  RAJ2000    DEJ2000        GLON  \\\n",
              "0  b'4FGL J0001.2+4741'            1   0.3126  47.685902  114.250198   \n",
              "1  b'4FGL J0001.2-0747'            1   0.3151  -7.797100   89.032722   \n",
              "2  b'4FGL J0001.4-0010'            3   0.3717  -0.169900   96.891991   \n",
              "3  b'4FGL J0001.5+2113'            1   0.3815  21.218300  107.649353   \n",
              "5  b'4FGL J0001.8-2153'            3   0.4647 -21.886499   55.184696   \n",
              "\n",
              "        GLAT  Signif_Avg      Flux1000  Unc_Flux1000  Energy_Flux100  ...  \\\n",
              "0 -14.338059    4.692600  1.259796e-10  2.877197e-11    1.499454e-12  ...   \n",
              "1 -67.305008   23.782486  7.471219e-10  5.041939e-11    8.339171e-12  ...   \n",
              "2 -60.491329    4.900018  1.082246e-10  2.914440e-11    1.231385e-12  ...   \n",
              "3 -40.167652   54.653084  1.347354e-09  6.172809e-11    2.555889e-11  ...   \n",
              "5 -77.572540    4.578068  9.047530e-11  2.442527e-11    1.061114e-12  ...   \n",
              "\n",
              "       HE_EPeak  Unc_HE_EPeak  HE_nuFnuPeak  Unc_HE_nuFnuPeak        nu_syn  \\\n",
              "0           NaN           NaN          -inf              -inf  1.000000e+14   \n",
              "1    796.297852    766.797424  8.686245e-07      7.492923e-08  9.120108e+13   \n",
              "2  15475.512695  18471.714844  1.494930e-07      8.294155e-08  3.609939e+12   \n",
              "3     65.680008     18.915735  8.844835e-06      8.229393e-07  1.584893e+14   \n",
              "5   6281.065430   2626.445068  1.545364e-07      5.278574e-08  1.659588e+13   \n",
              "\n",
              "      nuFnu_syn  Variability_Index  Frac_Variability  Unc_Frac_Variability  \\\n",
              "0  3.716456e-13          25.313953          0.675882              0.312850   \n",
              "1  1.936770e-12          46.780693          0.406565              0.111870   \n",
              "2  2.473256e-13           9.272764          0.000000             10.000000   \n",
              "3  1.152471e-12        1910.935791          0.996138              0.213478   \n",
              "5  9.716011e-13          24.557972          0.902851              0.380107   \n",
              "\n",
              "   Highest_energy  \n",
              "0       65.575203  \n",
              "1       86.975197  \n",
              "2       30.056200  \n",
              "3            -inf  \n",
              "5            -inf  \n",
              "\n",
              "[5 rows x 41 columns]"
            ]
          },
          "execution_count": 3,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_h.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 342
        },
        "id": "U2ZBJtwZ_Duq",
        "outputId": "52696738-c524-4779-8ea2-0de26f18fba5"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-e768954b-e1a2-4385-bedb-272be30af512\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Source_Name</th>\n",
              "      <th>DataRelease</th>\n",
              "      <th>RAJ2000</th>\n",
              "      <th>DEJ2000</th>\n",
              "      <th>GLON</th>\n",
              "      <th>GLAT</th>\n",
              "      <th>Signif_Avg</th>\n",
              "      <th>Flux1000</th>\n",
              "      <th>Unc_Flux1000</th>\n",
              "      <th>Energy_Flux100</th>\n",
              "      <th>...</th>\n",
              "      <th>HE_EPeak</th>\n",
              "      <th>Unc_HE_EPeak</th>\n",
              "      <th>HE_nuFnuPeak</th>\n",
              "      <th>Unc_HE_nuFnuPeak</th>\n",
              "      <th>nu_syn</th>\n",
              "      <th>nuFnu_syn</th>\n",
              "      <th>Variability_Index</th>\n",
              "      <th>Frac_Variability</th>\n",
              "      <th>Unc_Frac_Variability</th>\n",
              "      <th>Highest_energy</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>b'4FGL J0012.0+7043'</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0032</td>\n",
              "      <td>70.732597</td>\n",
              "      <td>119.661102</td>\n",
              "      <td>8.111637</td>\n",
              "      <td>9.391647</td>\n",
              "      <td>4.022406e-10</td>\n",
              "      <td>4.989643e-11</td>\n",
              "      <td>5.202780e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>-inf</td>\n",
              "      <td>-inf</td>\n",
              "      <td>5.754399e+12</td>\n",
              "      <td>5.482871e-13</td>\n",
              "      <td>18.997129</td>\n",
              "      <td>0.381762</td>\n",
              "      <td>0.192456</td>\n",
              "      <td>-inf</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>b'4FGL J0014.7+5801'</td>\n",
              "      <td>1</td>\n",
              "      <td>3.6867</td>\n",
              "      <td>58.027199</td>\n",
              "      <td>118.074745</td>\n",
              "      <td>-4.498106</td>\n",
              "      <td>10.829500</td>\n",
              "      <td>1.763550e-10</td>\n",
              "      <td>4.287831e-11</td>\n",
              "      <td>2.571515e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>30467.089844</td>\n",
              "      <td>17695.964844</td>\n",
              "      <td>5.071456e-07</td>\n",
              "      <td>1.101557e-07</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>11.195417</td>\n",
              "      <td>0.099530</td>\n",
              "      <td>0.650253</td>\n",
              "      <td>160.817001</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>b'4FGL J0014.8+6118'</td>\n",
              "      <td>1</td>\n",
              "      <td>3.7212</td>\n",
              "      <td>61.303699</td>\n",
              "      <td>118.557297</td>\n",
              "      <td>-1.257189</td>\n",
              "      <td>17.677473</td>\n",
              "      <td>7.848316e-10</td>\n",
              "      <td>7.590142e-11</td>\n",
              "      <td>6.474739e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>5067.017578</td>\n",
              "      <td>1501.950317</td>\n",
              "      <td>9.392241e-07</td>\n",
              "      <td>9.292127e-08</td>\n",
              "      <td>2.089296e+13</td>\n",
              "      <td>1.480430e-12</td>\n",
              "      <td>12.274307</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>10.000000</td>\n",
              "      <td>-inf</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>b'4FGL J0015.6+5551'</td>\n",
              "      <td>1</td>\n",
              "      <td>3.9071</td>\n",
              "      <td>55.863602</td>\n",
              "      <td>117.888100</td>\n",
              "      <td>-6.657064</td>\n",
              "      <td>13.679805</td>\n",
              "      <td>3.372334e-10</td>\n",
              "      <td>4.716959e-11</td>\n",
              "      <td>4.016745e-12</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>-inf</td>\n",
              "      <td>-inf</td>\n",
              "      <td>5.688530e+15</td>\n",
              "      <td>7.878908e-12</td>\n",
              "      <td>11.952563</td>\n",
              "      <td>0.136779</td>\n",
              "      <td>0.318575</td>\n",
              "      <td>167.578003</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>b'4FGL J0035.9+5950'</td>\n",
              "      <td>1</td>\n",
              "      <td>8.9823</td>\n",
              "      <td>59.833401</td>\n",
              "      <td>120.982048</td>\n",
              "      <td>-2.979620</td>\n",
              "      <td>68.426636</td>\n",
              "      <td>3.081422e-09</td>\n",
              "      <td>9.592344e-11</td>\n",
              "      <td>3.706086e-11</td>\n",
              "      <td>...</td>\n",
              "      <td>98809.250000</td>\n",
              "      <td>65159.152344</td>\n",
              "      <td>5.909368e-06</td>\n",
              "      <td>5.170865e-07</td>\n",
              "      <td>1.122018e+17</td>\n",
              "      <td>1.493479e-11</td>\n",
              "      <td>252.022751</td>\n",
              "      <td>0.494519</td>\n",
              "      <td>0.109907</td>\n",
              "      <td>372.923004</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 41 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e768954b-e1a2-4385-bedb-272be30af512')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-e768954b-e1a2-4385-bedb-272be30af512 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-e768954b-e1a2-4385-bedb-272be30af512');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-5f56a94e-73ec-4106-bd8d-7b39e628f642\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-5f56a94e-73ec-4106-bd8d-7b39e628f642')\"\n",
              "            title=\"Suggest charts.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "    background-color: #E8F0FE;\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: #1967D2;\n",
              "    height: 32px;\n",
              "    padding: 0 0 0 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: #E2EBFA;\n",
              "    box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: #174EA6;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "    background-color: #3B4455;\n",
              "    fill: #D2E3FC;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart:hover {\n",
              "    background-color: #434B5C;\n",
              "    box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "    filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "    fill: #FFFFFF;\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const charts = await google.colab.kernel.invokeFunction(\n",
              "          'suggestCharts', [key], {});\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-5f56a94e-73ec-4106-bd8d-7b39e628f642 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "            Source_Name  DataRelease  RAJ2000    DEJ2000        GLON  \\\n",
              "0  b'4FGL J0012.0+7043'            1   3.0032  70.732597  119.661102   \n",
              "1  b'4FGL J0014.7+5801'            1   3.6867  58.027199  118.074745   \n",
              "2  b'4FGL J0014.8+6118'            1   3.7212  61.303699  118.557297   \n",
              "3  b'4FGL J0015.6+5551'            1   3.9071  55.863602  117.888100   \n",
              "6  b'4FGL J0035.9+5950'            1   8.9823  59.833401  120.982048   \n",
              "\n",
              "       GLAT  Signif_Avg      Flux1000  Unc_Flux1000  Energy_Flux100  ...  \\\n",
              "0  8.111637    9.391647  4.022406e-10  4.989643e-11    5.202780e-12  ...   \n",
              "1 -4.498106   10.829500  1.763550e-10  4.287831e-11    2.571515e-12  ...   \n",
              "2 -1.257189   17.677473  7.848316e-10  7.590142e-11    6.474739e-12  ...   \n",
              "3 -6.657064   13.679805  3.372334e-10  4.716959e-11    4.016745e-12  ...   \n",
              "6 -2.979620   68.426636  3.081422e-09  9.592344e-11    3.706086e-11  ...   \n",
              "\n",
              "       HE_EPeak  Unc_HE_EPeak  HE_nuFnuPeak  Unc_HE_nuFnuPeak        nu_syn  \\\n",
              "0           NaN           NaN          -inf              -inf  5.754399e+12   \n",
              "1  30467.089844  17695.964844  5.071456e-07      1.101557e-07  0.000000e+00   \n",
              "2   5067.017578   1501.950317  9.392241e-07      9.292127e-08  2.089296e+13   \n",
              "3           NaN           NaN          -inf              -inf  5.688530e+15   \n",
              "6  98809.250000  65159.152344  5.909368e-06      5.170865e-07  1.122018e+17   \n",
              "\n",
              "      nuFnu_syn  Variability_Index  Frac_Variability  Unc_Frac_Variability  \\\n",
              "0  5.482871e-13          18.997129          0.381762              0.192456   \n",
              "1  0.000000e+00          11.195417          0.099530              0.650253   \n",
              "2  1.480430e-12          12.274307          0.000000             10.000000   \n",
              "3  7.878908e-12          11.952563          0.136779              0.318575   \n",
              "6  1.493479e-11         252.022751          0.494519              0.109907   \n",
              "\n",
              "   Highest_energy  \n",
              "0            -inf  \n",
              "1      160.817001  \n",
              "2            -inf  \n",
              "3      167.578003  \n",
              "6      372.923004  \n",
              "\n",
              "[5 rows x 41 columns]"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_l.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        },
        "id": "gNx-cog1_FUb",
        "outputId": "e12afacb-1496-4b9a-8657-f70c9dc36c09"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<>:8: SyntaxWarning: \"is not\" with a literal. Did you mean \"!=\"?\n",
            "<>:8: SyntaxWarning: \"is not\" with a literal. Did you mean \"!=\"?\n",
            "<ipython-input-5-6a33a8dc7af4>:8: SyntaxWarning: \"is not\" with a literal. Did you mean \"!=\"?\n",
            "  feature_columns = [ele for ele in selected_columns if ele is not 'CLASS']\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-23aa5e6a-53a2-4a1f-9368-5502f85ad5ca\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>PL_Index</th>\n",
              "      <th>nu_syn</th>\n",
              "      <th>LP_Index</th>\n",
              "      <th>Pivot_Energy</th>\n",
              "      <th>Frac_Variability</th>\n",
              "      <th>Variability_Index</th>\n",
              "      <th>nuFnu_syn</th>\n",
              "      <th>CLASS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2.271696</td>\n",
              "      <td>1.000000e+14</td>\n",
              "      <td>2.254081</td>\n",
              "      <td>2420.575195</td>\n",
              "      <td>0.675882</td>\n",
              "      <td>25.313953</td>\n",
              "      <td>3.716456e-13</td>\n",
              "      <td>BCU</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2.116692</td>\n",
              "      <td>9.120108e+13</td>\n",
              "      <td>2.078927</td>\n",
              "      <td>1721.603882</td>\n",
              "      <td>0.406565</td>\n",
              "      <td>46.780693</td>\n",
              "      <td>1.936770e-12</td>\n",
              "      <td>BLL</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.939160</td>\n",
              "      <td>3.609939e+12</td>\n",
              "      <td>1.661223</td>\n",
              "      <td>4307.082520</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>9.272764</td>\n",
              "      <td>2.473256e-13</td>\n",
              "      <td>BLL</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2.654060</td>\n",
              "      <td>1.584893e+14</td>\n",
              "      <td>2.514159</td>\n",
              "      <td>329.773743</td>\n",
              "      <td>0.996138</td>\n",
              "      <td>1910.935791</td>\n",
              "      <td>1.152471e-12</td>\n",
              "      <td>FSRQ</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.876663</td>\n",
              "      <td>1.659588e+13</td>\n",
              "      <td>1.716620</td>\n",
              "      <td>4429.934082</td>\n",
              "      <td>0.902851</td>\n",
              "      <td>24.557972</td>\n",
              "      <td>9.716011e-13</td>\n",
              "      <td>BCU</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3169</th>\n",
              "      <td>2.071567</td>\n",
              "      <td>3.311315e+13</td>\n",
              "      <td>2.042366</td>\n",
              "      <td>5897.854004</td>\n",
              "      <td>0.328630</td>\n",
              "      <td>14.602383</td>\n",
              "      <td>1.006315e-12</td>\n",
              "      <td>BCU</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3170</th>\n",
              "      <td>2.226648</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>1.962429</td>\n",
              "      <td>3114.226318</td>\n",
              "      <td>1.071415</td>\n",
              "      <td>34.235767</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>BCU</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3171</th>\n",
              "      <td>2.045913</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>2.019132</td>\n",
              "      <td>5066.525391</td>\n",
              "      <td>0.337927</td>\n",
              "      <td>22.359163</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>BCU</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3172</th>\n",
              "      <td>1.810977</td>\n",
              "      <td>1.566751e+16</td>\n",
              "      <td>1.736923</td>\n",
              "      <td>1910.885742</td>\n",
              "      <td>0.201833</td>\n",
              "      <td>58.688427</td>\n",
              "      <td>1.279994e-11</td>\n",
              "      <td>BLL</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3173</th>\n",
              "      <td>1.743453</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>1.572163</td>\n",
              "      <td>5740.775391</td>\n",
              "      <td>0.175460</td>\n",
              "      <td>13.854220</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>BCU</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3120 rows × 8 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-23aa5e6a-53a2-4a1f-9368-5502f85ad5ca')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-23aa5e6a-53a2-4a1f-9368-5502f85ad5ca button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-23aa5e6a-53a2-4a1f-9368-5502f85ad5ca');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-97506484-2a93-4866-a670-16b335074266\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-97506484-2a93-4866-a670-16b335074266')\"\n",
              "            title=\"Suggest charts.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "    background-color: #E8F0FE;\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: #1967D2;\n",
              "    height: 32px;\n",
              "    padding: 0 0 0 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: #E2EBFA;\n",
              "    box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: #174EA6;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "    background-color: #3B4455;\n",
              "    fill: #D2E3FC;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart:hover {\n",
              "    background-color: #434B5C;\n",
              "    box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "    filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "    fill: #FFFFFF;\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const charts = await google.colab.kernel.invokeFunction(\n",
              "          'suggestCharts', [key], {});\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-97506484-2a93-4866-a670-16b335074266 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "      PL_Index        nu_syn  LP_Index  Pivot_Energy  Frac_Variability  \\\n",
              "0     2.271696  1.000000e+14  2.254081   2420.575195          0.675882   \n",
              "1     2.116692  9.120108e+13  2.078927   1721.603882          0.406565   \n",
              "2     1.939160  3.609939e+12  1.661223   4307.082520          0.000000   \n",
              "3     2.654060  1.584893e+14  2.514159    329.773743          0.996138   \n",
              "4     1.876663  1.659588e+13  1.716620   4429.934082          0.902851   \n",
              "...        ...           ...       ...           ...               ...   \n",
              "3169  2.071567  3.311315e+13  2.042366   5897.854004          0.328630   \n",
              "3170  2.226648  0.000000e+00  1.962429   3114.226318          1.071415   \n",
              "3171  2.045913  0.000000e+00  2.019132   5066.525391          0.337927   \n",
              "3172  1.810977  1.566751e+16  1.736923   1910.885742          0.201833   \n",
              "3173  1.743453  0.000000e+00  1.572163   5740.775391          0.175460   \n",
              "\n",
              "      Variability_Index     nuFnu_syn CLASS  \n",
              "0             25.313953  3.716456e-13   BCU  \n",
              "1             46.780693  1.936770e-12   BLL  \n",
              "2              9.272764  2.473256e-13   BLL  \n",
              "3           1910.935791  1.152471e-12  FSRQ  \n",
              "4             24.557972  9.716011e-13   BCU  \n",
              "...                 ...           ...   ...  \n",
              "3169          14.602383  1.006315e-12   BCU  \n",
              "3170          34.235767  0.000000e+00   BCU  \n",
              "3171          22.359163  0.000000e+00   BCU  \n",
              "3172          58.688427  1.279994e-11   BLL  \n",
              "3173          13.854220  0.000000e+00   BCU  \n",
              "\n",
              "[3120 rows x 8 columns]"
            ]
          },
          "execution_count": 5,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "classification_df = pd.concat([df_h, df_l], ignore_index=True)\n",
        "classification_df['CLASS'] = classification_df['CLASS'].str.decode('utf-8')\n",
        "classification_df['CLASS'] = classification_df['CLASS'].str.upper()\n",
        "classification_df = classification_df[classification_df['CLASS'].isin(['BLL', 'FSRQ', 'BCU'])]\n",
        "\n",
        "selected_columns = ['PL_Index', 'nu_syn', 'LP_Index', 'Pivot_Energy', 'Frac_Variability', 'Variability_Index', 'nuFnu_syn', 'CLASS']\n",
        "target = 'CLASS'\n",
        "feature_columns = [ele for ele in selected_columns if ele is not 'CLASS']\n",
        "\n",
        "\n",
        "classification_df = classification_df[selected_columns]\n",
        "\n",
        "classification_df"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iiuGs12at3wZ",
        "outputId": "fc723134-73aa-4d5f-a7ad-0aed5c8d4dd2"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "(2005, 8)\n",
            "(1115, 8)\n"
          ]
        }
      ],
      "source": [
        "bcu_samples = classification_df[classification_df['CLASS'].isin(['BCU'])]\n",
        "classification_df = classification_df[~classification_df['CLASS'].isin(['BCU'])]\n",
        "\n",
        "print(classification_df.shape)\n",
        "print(bcu_samples.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3QVP23iMG3YA",
        "outputId": "9e5a40cb-c8fd-4c51-f371-9463d63ef948"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-9-23a26e19f24f>:2: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  classification_df['CLASS'] = classification_df['CLASS'].map(class_mapping)\n"
          ]
        }
      ],
      "source": [
        "class_mapping = {'BLL': 0, 'FSRQ': 1}\n",
        "classification_df['CLASS'] = classification_df['CLASS'].map(class_mapping)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "7XIFt4QfsKJ0"
      },
      "outputs": [],
      "source": [
        "from sklearn.compose import make_column_transformer\n",
        "from sklearn.preprocessing import OneHotEncoder, StandardScaler, LabelEncoder, OrdinalEncoder"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "qAV3e2nsru6J"
      },
      "outputs": [],
      "source": [
        "transformer = make_column_transformer(\n",
        "    (StandardScaler(),\n",
        "        feature_columns)\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "m26Mxtd8rwpJ"
      },
      "outputs": [],
      "source": [
        "features = classification_df.drop([target], axis=1)\n",
        "labels = classification_df[target]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "RgmkFNjNsoG0"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import train_test_split"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "7tgpQZGssdfI"
      },
      "outputs": [],
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.20, random_state=42)\n",
        "X_val, X_test, y_val, y_test = train_test_split(X_test, y_test, test_size=0.50, random_state=42)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "y3N2oVjYsgEx"
      },
      "outputs": [],
      "source": [
        "X_train.to_csv('train_samples.csv', index=False)\n",
        "X_val.to_csv('val_samples.csv', index = False)\n",
        "X_test.to_csv('test_samples.csv', index = False)\n",
        "y_train.to_csv('train_labels.csv', index = False)\n",
        "y_test.to_csv('test_labels.csv', index = False)\n",
        "y_val.to_csv('val_labels.csv', index = False)\n",
        "bcu_samples.to_csv('bcu_samples.csv', index = False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "PmF1EeePsuSA"
      },
      "outputs": [],
      "source": [
        "transformer.fit(X_train)\n",
        "X_train = transformer.transform(X_train)\n",
        "X_test = transformer.transform(X_test)\n",
        "X_val = transformer.transform(X_val)\n",
        "bcu_samples = transformer.transform(bcu_samples)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "kgrNrVKfuXi8"
      },
      "outputs": [],
      "source": [
        "np.save('X_train.npy', X_train)\n",
        "np.save('X_val.npy', X_val)\n",
        "np.save('X_test.npy', X_test)\n",
        "np.save('y_train.npy', y_train)\n",
        "np.save('y_val.npy', y_val)\n",
        "np.save('y_test.npy', y_test)\n",
        "np.save('bcu_samples.npy', bcu_samples)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "w-S7vXhXuvtu"
      },
      "outputs": [],
      "source": [
        "X_train = np.load('X_train.npy')\n",
        "X_val = np.load('X_val.npy')\n",
        "X_test = np.load('X_test.npy')\n",
        "y_train = np.load('y_train.npy')\n",
        "y_val = np.load('y_val.npy')\n",
        "y_test = np.load('y_test.npy')\n",
        "bcu_samples = np.load('bcu_samples.npy')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {},
      "outputs": [],
      "source": [
        "X_train = np.load('./classification_data/X_train.npy')\n",
        "X_val = np.load('./classification_data/X_val.npy')\n",
        "X_test = np.load('./classification_data/X_test.npy')\n",
        "y_train = np.load('./classification_data/y_train.npy')\n",
        "y_val = np.load('./classification_data/y_val.npy')\n",
        "y_test = np.load('./classification_data/y_test.npy')\n",
        "bcu_samples = np.load('./classification_data/bcu_samples.npy')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "L0r-dqWnu89G",
        "outputId": "538c5f84-1bd3-4d41-8867-8c6013eb1a16"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "1053\n",
            "551\n"
          ]
        }
      ],
      "source": [
        "bll, fsrq = np.bincount(y_train)\n",
        "print(bll)\n",
        "print(fsrq)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FSEYWqGou-hW",
        "outputId": "27a44434-019e-41f5-94a1-3ec083bc7507"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "((1604, 7), (1604,))"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "X_train.shape,y_train.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1yKaNqntvAeR",
        "outputId": "7038c0b4-88ba-4761-80f8-acd36b81d985"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([ 0.85295165, -0.09814737,  0.71171683, -0.7151801 , -0.27865052,\n",
              "       -0.17254569,  0.11640763], dtype=float32)"
            ]
          },
          "execution_count": 21,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "X_train[0]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "peaoICdM_MUr"
      },
      "source": [
        "## Autoencoder Pretext"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EXyzUQDYvC8o",
        "outputId": "9e5c8633-33f2-42a7-e846-66e119a6bee8"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1200x800 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "accuracy\n",
            "\ttraining         \t (min:    0.549, max:    0.908, cur:    0.907)\n",
            "\tvalidation       \t (min:    0.585, max:    0.880, cur:    0.880)\n",
            "Loss\n",
            "\ttraining         \t (min:    0.253, max:    0.686, cur:    0.253)\n",
            "\tvalidation       \t (min:    0.265, max:    0.661, cur:    0.265)\n",
            "\n",
            "Epoch 01000: saving model to ./ssl_ckpt_no_dropout\\weights_epoch_1000.h5\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x1c0136dddc8>"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras.layers import Input, Dense\n",
        "from keras.models import Model\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "\n",
        "# Define the self-supervised model architecture\n",
        "input_layer = Input(shape=(7,))\n",
        "#encoder = Dense(128, activation='relu')(input_layer)\n",
        "encoder = Dense(64, activation='relu')(input_layer)\n",
        "encoded = Dense(32, activation='relu')(encoder)\n",
        "decoder = Dense(64, activation='relu')(encoded)\n",
        "#decoder = Dense(128, activation='relu')(decoder)\n",
        "decoded = Dense(7, activation='linear')(decoder)  # Reconstruction layer\n",
        "\n",
        "# Create the encoder model (used for self-supervised task)\n",
        "encoder_model = Model(inputs=input_layer, outputs=encoded)\n",
        "\n",
        "# Create the full self-supervised model (including reconstruction loss)\n",
        "self_supervised_model = Model(inputs=input_layer, outputs=decoded)\n",
        "\n",
        "# Compile the self-supervised model with a loss function (reconstruction loss)\n",
        "self_supervised_model.compile(optimizer='adam', loss='mean_squared_error')\n",
        "\n",
        "# Train the self-supervised model on the data\n",
        "self_supervised_model.fit(X_train, X_train, epochs=300, batch_size=2048, validation_data=(X_val, X_val), callbacks=[PlotLossesKeras()])\n",
        "\n",
        "\n",
        "encoder_model.save_weights('./ssl_encoder.h5')\n",
        "# Extract encoded features from the encoder model\n",
        "encoded_features_train = encoder_model.predict(X_train)\n",
        "encoded_features_val = encoder_model.predict(X_val)\n",
        "\n",
        "# Now you can use the encoded features as inputs for a classification head\n",
        "classification_input_layer = Input(shape=(32,))  # Use the appropriate size of encoded features\n",
        "#classification_input_layer_1 =  tf.keras.layers.Dropout(0.25)(classification_input_layer)\n",
        "classification_output_layer = Dense(1, activation='sigmoid')(classification_input_layer)\n",
        "\n",
        "classification_model = Model(inputs=classification_input_layer, outputs=classification_output_layer)\n",
        "classification_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "checkpoint_path = './ssl_ckpt_no_dropout/' + 'weights_epoch_{epoch:02d}.h5'\n",
        "ckpt = tf.keras.callbacks.ModelCheckpoint(\n",
        "        checkpoint_path,\n",
        "        monitor='val_loss',\n",
        "        verbose=1,\n",
        "        save_best_only=False,\n",
        "        save_weights_only=True,\n",
        "        mode='min',\n",
        "        save_freq='epoch'\n",
        "    )\n",
        "\n",
        "# Train the classification model on the encoded features\n",
        "classification_model.fit(encoded_features_train, y_train, epochs=1000, batch_size=2048, validation_data=(encoded_features_val, y_val), callbacks=[PlotLossesKeras(), ckpt])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 67%|██████▋   | 1000/1500 [00:57<00:29, 16.70it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 67%|██████▋   | 1002/1500 [00:57<00:31, 16.05it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 67%|██████▋   | 1004/1500 [00:57<00:31, 15.80it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 67%|██████▋   | 1006/1500 [00:57<00:31, 15.71it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 67%|██████▋   | 1008/1500 [00:58<00:31, 15.81it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 67%|██████▋   | 1010/1500 [00:58<00:30, 15.90it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 67%|██████▋   | 1012/1500 [00:58<00:29, 16.50it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 68%|██████▊   | 1014/1500 [00:58<00:28, 16.99it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 68%|██████▊   | 1016/1500 [00:58<00:29, 16.19it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 68%|██████▊   | 1018/1500 [00:58<00:31, 15.54it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 68%|██████▊   | 1020/1500 [00:58<00:30, 15.55it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 68%|██████▊   | 1022/1500 [00:58<00:30, 15.88it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 68%|██████▊   | 1024/1500 [00:59<00:29, 15.91it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 68%|██████▊   | 1026/1500 [00:59<00:28, 16.56it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▊   | 1028/1500 [00:59<00:28, 16.54it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▊   | 1030/1500 [00:59<00:29, 15.93it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▉   | 1032/1500 [00:59<00:28, 16.57it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▉   | 1034/1500 [00:59<00:28, 16.10it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▉   | 1036/1500 [00:59<00:28, 16.07it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▉   | 1038/1500 [00:59<00:27, 16.68it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▉   | 1040/1500 [01:00<00:27, 16.47it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 69%|██████▉   | 1042/1500 [01:00<00:26, 16.97it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 70%|██████▉   | 1044/1500 [01:00<00:28, 15.86it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 70%|██████▉   | 1046/1500 [01:00<00:27, 16.56it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 70%|██████▉   | 1048/1500 [01:00<00:28, 15.97it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 70%|███████   | 1050/1500 [01:00<00:29, 15.29it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 70%|███████   | 1052/1500 [01:00<00:28, 15.62it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 70%|███████   | 1054/1500 [01:00<00:29, 15.08it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 70%|███████   | 1056/1500 [01:01<00:28, 15.75it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████   | 1058/1500 [01:01<00:29, 15.00it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████   | 1060/1500 [01:01<00:29, 15.16it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████   | 1062/1500 [01:01<00:28, 15.53it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████   | 1064/1500 [01:01<00:28, 15.21it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████   | 1066/1500 [01:01<00:29, 14.84it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████   | 1068/1500 [01:01<00:29, 14.72it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████▏  | 1070/1500 [01:01<00:28, 15.15it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 71%|███████▏  | 1072/1500 [01:02<00:28, 15.02it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 72%|███████▏  | 1074/1500 [01:02<00:29, 14.48it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 72%|███████▏  | 1076/1500 [01:02<00:30, 14.12it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 72%|███████▏  | 1078/1500 [01:02<00:27, 15.16it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 72%|███████▏  | 1080/1500 [01:02<00:27, 15.29it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 72%|███████▏  | 1082/1500 [01:02<00:26, 15.76it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 72%|███████▏  | 1084/1500 [01:02<00:26, 15.48it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 72%|███████▏  | 1086/1500 [01:03<00:27, 15.13it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1088/1500 [01:03<00:26, 15.63it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1090/1500 [01:03<00:26, 15.49it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1092/1500 [01:03<00:26, 15.24it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1094/1500 [01:03<00:26, 15.45it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1096/1500 [01:03<00:25, 16.08it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1098/1500 [01:03<00:25, 15.72it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1100/1500 [01:03<00:25, 15.94it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 73%|███████▎  | 1102/1500 [01:04<00:25, 15.75it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 74%|███████▎  | 1104/1500 [01:04<00:23, 16.52it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 74%|███████▎  | 1106/1500 [01:04<00:23, 17.01it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 74%|███████▍  | 1108/1500 [01:04<00:23, 16.69it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 74%|███████▍  | 1110/1500 [01:04<00:22, 17.14it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 74%|███████▍  | 1112/1500 [01:04<00:23, 16.59it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 74%|███████▍  | 1114/1500 [01:04<00:23, 16.33it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 74%|███████▍  | 1116/1500 [01:04<00:23, 16.40it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▍  | 1118/1500 [01:05<00:22, 17.02it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▍  | 1120/1500 [01:05<00:22, 16.70it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▍  | 1122/1500 [01:05<00:24, 15.46it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▍  | 1124/1500 [01:05<00:24, 15.57it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▌  | 1126/1500 [01:05<00:24, 15.20it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▌  | 1128/1500 [01:05<00:25, 14.80it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▌  | 1130/1500 [01:05<00:25, 14.69it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 75%|███████▌  | 1132/1500 [01:06<00:27, 13.31it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 76%|███████▌  | 1134/1500 [01:06<00:27, 13.12it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 76%|███████▌  | 1136/1500 [01:06<00:27, 13.16it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 76%|███████▌  | 1138/1500 [01:06<00:27, 13.40it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 76%|███████▌  | 1140/1500 [01:06<00:37,  9.53it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 76%|███████▌  | 1142/1500 [01:06<00:33, 10.82it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 76%|███████▋  | 1144/1500 [01:07<00:30, 11.72it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 76%|███████▋  | 1146/1500 [01:07<00:27, 12.76it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1148/1500 [01:07<00:27, 12.70it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1150/1500 [01:07<00:26, 13.45it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1152/1500 [01:07<00:24, 14.06it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1154/1500 [01:07<00:24, 14.36it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1156/1500 [01:07<00:23, 14.65it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1158/1500 [01:07<00:22, 15.39it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1160/1500 [01:08<00:22, 15.32it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 77%|███████▋  | 1162/1500 [01:08<00:23, 14.10it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 78%|███████▊  | 1164/1500 [01:08<00:24, 13.70it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 78%|███████▊  | 1166/1500 [01:08<00:23, 13.99it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 78%|███████▊  | 1168/1500 [01:08<00:24, 13.76it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 78%|███████▊  | 1170/1500 [01:08<00:23, 13.82it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 78%|███████▊  | 1172/1500 [01:09<00:23, 13.93it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 78%|███████▊  | 1174/1500 [01:09<00:23, 14.01it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 78%|███████▊  | 1176/1500 [01:09<00:22, 14.13it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▊  | 1178/1500 [01:09<00:22, 14.31it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▊  | 1180/1500 [01:09<00:22, 14.51it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▉  | 1182/1500 [01:09<00:21, 14.48it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▉  | 1184/1500 [01:09<00:22, 14.20it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▉  | 1186/1500 [01:10<00:22, 13.88it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▉  | 1188/1500 [01:10<00:23, 13.04it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▉  | 1190/1500 [01:10<00:24, 12.90it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 79%|███████▉  | 1192/1500 [01:10<00:23, 12.96it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 80%|███████▉  | 1194/1500 [01:10<00:23, 13.18it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 80%|███████▉  | 1196/1500 [01:10<00:22, 13.23it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 80%|███████▉  | 1198/1500 [01:10<00:22, 13.55it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 80%|████████  | 1200/1500 [01:11<00:22, 13.39it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 80%|████████  | 1202/1500 [01:11<00:21, 13.74it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 80%|████████  | 1204/1500 [01:11<00:21, 13.90it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 80%|████████  | 1206/1500 [01:11<00:21, 13.93it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████  | 1208/1500 [01:11<00:21, 13.60it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████  | 1210/1500 [01:11<00:22, 12.96it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████  | 1212/1500 [01:11<00:22, 12.95it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████  | 1214/1500 [01:12<00:21, 13.47it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████  | 1216/1500 [01:12<00:20, 13.90it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████  | 1218/1500 [01:12<00:20, 13.98it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████▏ | 1220/1500 [01:12<00:18, 14.77it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 81%|████████▏ | 1222/1500 [01:12<00:19, 14.40it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 82%|████████▏ | 1224/1500 [01:12<00:18, 14.67it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 82%|████████▏ | 1226/1500 [01:12<00:18, 14.72it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 82%|████████▏ | 1228/1500 [01:13<00:20, 13.46it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 82%|████████▏ | 1230/1500 [01:13<00:19, 13.84it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 82%|████████▏ | 1232/1500 [01:13<00:19, 13.64it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 82%|████████▏ | 1234/1500 [01:13<00:18, 14.12it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 82%|████████▏ | 1236/1500 [01:13<00:18, 14.63it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1238/1500 [01:13<00:18, 14.56it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1240/1500 [01:13<00:16, 15.51it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1242/1500 [01:13<00:15, 16.19it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1244/1500 [01:14<00:16, 15.56it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1246/1500 [01:14<00:15, 16.05it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1248/1500 [01:14<00:15, 16.03it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1250/1500 [01:14<00:15, 16.65it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 83%|████████▎ | 1252/1500 [01:14<00:15, 16.52it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 84%|████████▎ | 1254/1500 [01:14<00:16, 15.20it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 84%|████████▎ | 1256/1500 [01:14<00:15, 15.28it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 84%|████████▍ | 1258/1500 [01:15<00:16, 14.84it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 84%|████████▍ | 1260/1500 [01:15<00:16, 14.26it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 84%|████████▍ | 1262/1500 [01:15<00:16, 14.24it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 84%|████████▍ | 1264/1500 [01:15<00:18, 12.88it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 84%|████████▍ | 1266/1500 [01:15<00:19, 12.29it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▍ | 1268/1500 [01:15<00:18, 12.41it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▍ | 1270/1500 [01:15<00:17, 12.99it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▍ | 1272/1500 [01:16<00:16, 13.51it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▍ | 1274/1500 [01:16<00:16, 14.07it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▌ | 1276/1500 [01:16<00:17, 13.17it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▌ | 1278/1500 [01:16<00:16, 13.70it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▌ | 1280/1500 [01:16<00:15, 13.82it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 85%|████████▌ | 1282/1500 [01:16<00:15, 13.91it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 86%|████████▌ | 1284/1500 [01:16<00:15, 14.05it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 86%|████████▌ | 1286/1500 [01:17<00:14, 14.61it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 86%|████████▌ | 1288/1500 [01:17<00:14, 14.93it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 86%|████████▌ | 1290/1500 [01:17<00:13, 15.27it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 86%|████████▌ | 1292/1500 [01:17<00:13, 15.27it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 86%|████████▋ | 1294/1500 [01:17<00:13, 15.37it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 86%|████████▋ | 1296/1500 [01:17<00:13, 15.27it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1298/1500 [01:17<00:13, 15.49it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1300/1500 [01:18<00:13, 15.30it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1302/1500 [01:18<00:13, 15.12it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1304/1500 [01:18<00:12, 15.33it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1306/1500 [01:18<00:13, 14.60it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1308/1500 [01:18<00:14, 12.95it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1310/1500 [01:18<00:14, 13.22it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 87%|████████▋ | 1312/1500 [01:18<00:13, 13.78it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 88%|████████▊ | 1314/1500 [01:19<00:13, 14.09it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 88%|████████▊ | 1316/1500 [01:19<00:12, 14.31it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 88%|████████▊ | 1318/1500 [01:19<00:12, 14.65it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 88%|████████▊ | 1320/1500 [01:19<00:12, 14.98it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 88%|████████▊ | 1322/1500 [01:19<00:12, 14.32it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 88%|████████▊ | 1324/1500 [01:19<00:12, 14.29it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 88%|████████▊ | 1326/1500 [01:19<00:12, 14.14it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▊ | 1328/1500 [01:19<00:12, 14.31it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▊ | 1330/1500 [01:20<00:11, 14.30it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▉ | 1332/1500 [01:20<00:11, 14.06it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▉ | 1334/1500 [01:20<00:11, 14.26it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▉ | 1336/1500 [01:20<00:11, 13.78it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▉ | 1338/1500 [01:20<00:11, 14.04it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▉ | 1340/1500 [01:20<00:11, 14.08it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 89%|████████▉ | 1342/1500 [01:20<00:11, 14.27it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 90%|████████▉ | 1344/1500 [01:21<00:10, 14.49it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 90%|████████▉ | 1346/1500 [01:21<00:10, 14.25it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 90%|████████▉ | 1348/1500 [01:21<00:10, 14.27it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 90%|█████████ | 1350/1500 [01:21<00:10, 14.28it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 90%|█████████ | 1352/1500 [01:21<00:10, 14.34it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 90%|█████████ | 1354/1500 [01:21<00:09, 14.60it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 90%|█████████ | 1356/1500 [01:22<00:10, 13.30it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████ | 1358/1500 [01:22<00:10, 13.35it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████ | 1360/1500 [01:22<00:10, 13.10it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████ | 1362/1500 [01:22<00:10, 12.85it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████ | 1364/1500 [01:22<00:10, 13.15it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████ | 1366/1500 [01:22<00:09, 13.55it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████ | 1368/1500 [01:22<00:09, 13.81it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████▏| 1370/1500 [01:23<00:09, 13.44it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 91%|█████████▏| 1372/1500 [01:23<00:09, 13.06it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 92%|█████████▏| 1374/1500 [01:23<00:09, 12.95it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 92%|█████████▏| 1376/1500 [01:23<00:09, 13.45it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 92%|█████████▏| 1378/1500 [01:23<00:08, 13.85it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 92%|█████████▏| 1380/1500 [01:23<00:08, 13.78it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 92%|█████████▏| 1382/1500 [01:23<00:08, 13.88it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 92%|█████████▏| 1384/1500 [01:24<00:08, 14.06it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 92%|█████████▏| 1386/1500 [01:24<00:07, 14.38it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1388/1500 [01:24<00:08, 13.20it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1390/1500 [01:24<00:08, 13.45it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1392/1500 [01:24<00:07, 13.76it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1394/1500 [01:24<00:07, 14.16it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1396/1500 [01:24<00:07, 13.61it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1398/1500 [01:25<00:07, 13.59it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1400/1500 [01:25<00:07, 13.82it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 93%|█████████▎| 1402/1500 [01:25<00:06, 14.23it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 94%|█████████▎| 1404/1500 [01:25<00:06, 14.18it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 94%|█████████▎| 1406/1500 [01:25<00:06, 14.27it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 94%|█████████▍| 1408/1500 [01:25<00:06, 13.63it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 94%|█████████▍| 1410/1500 [01:25<00:06, 13.88it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 94%|█████████▍| 1412/1500 [01:26<00:06, 14.09it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 94%|█████████▍| 1414/1500 [01:26<00:06, 13.90it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 94%|█████████▍| 1416/1500 [01:26<00:06, 13.22it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▍| 1418/1500 [01:26<00:06, 11.75it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▍| 1420/1500 [01:26<00:06, 11.75it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▍| 1422/1500 [01:26<00:06, 11.79it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▍| 1424/1500 [01:27<00:06, 12.35it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▌| 1426/1500 [01:27<00:05, 12.83it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▌| 1428/1500 [01:27<00:05, 13.11it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▌| 1430/1500 [01:27<00:05, 13.73it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 95%|█████████▌| 1432/1500 [01:27<00:05, 13.55it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 96%|█████████▌| 1434/1500 [01:27<00:04, 13.79it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 96%|█████████▌| 1436/1500 [01:27<00:04, 13.72it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 96%|█████████▌| 1438/1500 [01:28<00:04, 13.94it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 96%|█████████▌| 1440/1500 [01:28<00:04, 14.22it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 96%|█████████▌| 1442/1500 [01:28<00:04, 14.42it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 96%|█████████▋| 1444/1500 [01:28<00:04, 13.56it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 96%|█████████▋| 1446/1500 [01:28<00:04, 13.10it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1448/1500 [01:28<00:04, 12.86it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1450/1500 [01:29<00:03, 12.52it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1452/1500 [01:29<00:04, 11.53it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1454/1500 [01:29<00:03, 12.04it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1456/1500 [01:29<00:03, 12.51it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1458/1500 [01:29<00:03, 13.12it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1460/1500 [01:29<00:02, 13.77it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 97%|█████████▋| 1462/1500 [01:29<00:02, 14.25it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 98%|█████████▊| 1464/1500 [01:30<00:02, 14.57it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 98%|█████████▊| 1466/1500 [01:30<00:02, 14.85it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 98%|█████████▊| 1468/1500 [01:30<00:02, 15.21it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 98%|█████████▊| 1470/1500 [01:30<00:01, 15.59it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 98%|█████████▊| 1472/1500 [01:30<00:01, 15.84it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 98%|█████████▊| 1474/1500 [01:30<00:01, 15.33it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 98%|█████████▊| 1476/1500 [01:30<00:01, 14.99it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▊| 1478/1500 [01:30<00:01, 14.92it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▊| 1480/1500 [01:31<00:01, 13.97it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▉| 1482/1500 [01:31<00:01, 13.92it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▉| 1484/1500 [01:31<00:01, 13.64it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▉| 1486/1500 [01:31<00:01, 13.60it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▉| 1488/1500 [01:31<00:00, 13.72it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▉| 1490/1500 [01:31<00:00, 14.08it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            " 99%|█████████▉| 1492/1500 [01:31<00:00, 13.98it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "100%|█████████▉| 1494/1500 [01:32<00:00, 14.33it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "100%|█████████▉| 1496/1500 [01:32<00:00, 14.42it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "100%|█████████▉| 1498/1500 [01:32<00:00, 14.80it/s]c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "c:\\Users\\abhim\\anaconda3\\envs\\sdss\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1318: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "100%|██████████| 1500/1500 [01:32<00:00, 16.21it/s]"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Best ROC AUC: 0.95\n",
            "Best accuracy: 0.89\n",
            "Best PR(0): 0.94\n",
            "Best R(0): 1.0\n",
            "Best F1(0): 0.92\n",
            "Best PR(1): 0.78\n",
            "Best R(1): 0.86\n",
            "Best F1(1): 0.82\n",
            "Best F1: 0.89\n",
            "Best precision: 0.9\n",
            "Best recall: 0.89\n",
            "{'auc': './ssl_ckpt_no_dropout/weights_epoch_990.h5', 'accuracy': './ssl_ckpt_no_dropout/weights_epoch_680.h5', 'f1': './ssl_ckpt_no_dropout/weights_epoch_680.h5', 'f1_0': './ssl_ckpt_no_dropout/weights_epoch_680.h5', 'f1_1': './ssl_ckpt_no_dropout/weights_epoch_680.h5', 'precision': './ssl_ckpt_no_dropout/weights_epoch_680.h5', 'r0': './ssl_ckpt_no_dropout/weights_epoch_1001.h5', 'r1': './ssl_ckpt_no_dropout/weights_epoch_648.h5', 'pr_0': './ssl_ckpt_no_dropout/weights_epoch_680.h5', 'pr_1': './ssl_ckpt_no_dropout/weights_epoch_680.h5', 'recall': './ssl_ckpt_no_dropout/weights_epoch_680.h5'}\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import roc_auc_score, classification_report, accuracy_score, f1_score, precision_score, recall_score\n",
        "from tqdm import tqdm\n",
        "best_auc = 0\n",
        "best_accuracy = 0\n",
        "best_f1 = 0\n",
        "best_f1_0 = 0\n",
        "best_f1_1 = 0\n",
        "best_precision = 0\n",
        "best_pr_0 = 0\n",
        "best_pr_1 = 0\n",
        "best_recall = 0\n",
        "best_r0 = 0\n",
        "best_r1 = 0\n",
        "best_weights = {\n",
        "    'auc': None, \n",
        "    'accuracy': None, \n",
        "    'f1':None, \n",
        "    'f1_0':None, \n",
        "    'f1_1':None,\n",
        "    'precision': None,\n",
        "    'r0': None,\n",
        "    'r1': None,\n",
        "    'pr_0':None,\n",
        "    'pr_1':None,\n",
        "    'recall': None\n",
        "}\n",
        "encoded_features_test = encoder_model.predict(X_test, verbose=0)\n",
        "for epoch in tqdm(range(len(os.listdir('./ssl_ckpt_no_dropout/')))):\n",
        "    weights_path = f\"./ssl_ckpt_no_dropout/weights_epoch_{epoch+1:02d}.h5\"\n",
        "    classification_model.load_weights(weights_path)\n",
        "    \n",
        "    # Evaluate the model\n",
        "    predictions = classification_model.predict(encoded_features_test,verbose=0)\n",
        "    roc_auc = roc_auc_score(y_test, predictions)\n",
        "    accuracy = accuracy_score(y_test, np.round(predictions))\n",
        "    f1_0 = f1_score(y_test, np.round(predictions), pos_label=0)\n",
        "    f1_1 = f1_score(y_test, np.round(predictions), pos_label=1)\n",
        "    f1 = f1_score(y_test, np.round(predictions), average='weighted')\n",
        "    precision = precision_score(y_test, np.round(predictions), average='weighted')\n",
        "    pr_0 = precision_score(y_test, np.round(predictions), pos_label=0)\n",
        "    pr_1 = precision_score(y_test, np.round(predictions), pos_label=1)\n",
        "    recall = recall_score(y_test, np.round(predictions), average='weighted')\n",
        "    r0 = recall_score(y_test, np.round(predictions), pos_label=0)\n",
        "    r1 = recall_score(y_test, np.round(predictions), pos_label=1)\n",
        "    \n",
        "    #accuracy = standard_nn_classifier.evaluate(X_test, y_test, verbose=0)[1]\n",
        "    \n",
        "    if roc_auc > best_auc:\n",
        "        best_auc = roc_auc\n",
        "        best_weights['auc'] = weights_path\n",
        "    if accuracy > best_accuracy:\n",
        "        best_accuracy = accuracy\n",
        "        best_weights['accuracy'] = weights_path\n",
        "    if f1 > best_f1:\n",
        "        best_f1 = f1\n",
        "        best_weights['f1'] = weights_path\n",
        "    if f1_0 > best_f1_0:\n",
        "        best_f1_0 = f1_0\n",
        "        best_weights['f1_0'] = weights_path\n",
        "    if f1_1 > best_f1_1:\n",
        "        best_f1_1 = f1_1\n",
        "        best_weights['f1_1'] = weights_path\n",
        "    if precision > best_precision:\n",
        "        best_precision = precision\n",
        "        best_weights['precision'] = weights_path\n",
        "    if pr_0 > best_pr_0:\n",
        "        best_pr_0 = pr_0 \n",
        "        best_weights['pr_0'] = weights_path\n",
        "    if pr_1 > best_pr_1:\n",
        "        best_pr_1 = pr_1 \n",
        "        best_weights['pr_1'] = weights_path\n",
        "    if r0 > best_r0:\n",
        "        best_r0 = r0 \n",
        "        best_weights['r0'] = weights_path\n",
        "    if r1 > best_r1:\n",
        "        best_r1 = r1 \n",
        "        best_weights['r1'] = weights_path\n",
        "    if recall > best_recall:\n",
        "        best_recall = recall\n",
        "        best_weights['recall'] = weights_path\n",
        "\n",
        "print(\"Best ROC AUC:\", round(best_auc,2))\n",
        "print(\"Best accuracy:\", round(best_accuracy,2))\n",
        "print(\"Best PR(0):\", round(best_pr_0,2))\n",
        "print(\"Best R(0):\", round(best_r0,2))\n",
        "print(\"Best F1(0):\", round(best_f1_0,2))\n",
        "print(\"Best PR(1):\", round(best_pr_1,2))\n",
        "print(\"Best R(1):\", round(best_r1,2))\n",
        "print(\"Best F1(1):\", round(best_f1_1,2))\n",
        "print(\"Best F1:\", round(best_f1,2))\n",
        "print('Best precision:', round(best_precision,2))\n",
        "print('Best recall:', round(best_recall,2))\n",
        "print(best_weights)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0.9491525423728814\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0      0.941     0.901     0.921       142\n",
            "           1      0.785     0.864     0.823        59\n",
            "\n",
            "    accuracy                          0.891       201\n",
            "   macro avg      0.863     0.883     0.872       201\n",
            "weighted avg      0.895     0.891     0.892       201\n",
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import roc_auc_score, classification_report\n",
        "classification_model.load_weights('./ssl_ckpt_no_dropout/weights_epoch_680.h5')\n",
        "predictions = classification_model.predict(encoded_features_test)\n",
        "print(roc_auc_score(y_test, predictions))\n",
        "print(classification_report(y_test, np.round(predictions), digits = 3))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sns.set(style=\"dark\")\n",
        "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
        "cm = confusion_matrix(y_test, np.round(predictions))\n",
        "disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=['BLL', 'FSRQ'])\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1200x800 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "accuracy\n",
            "\ttraining         \t (min:    0.301, max:    0.915, cur:    0.914)\n",
            "\tvalidation       \t (min:    0.275, max:    0.910, cur:    0.905)\n",
            "Loss\n",
            "\ttraining         \t (min:    0.253, max:    1.101, cur:    0.253)\n",
            "\tvalidation       \t (min:    0.251, max:    1.196, cur:    0.251)\n",
            "\n",
            "Epoch 01500: saving model to ./ssl_ckpt_no_dropout\\weights_epoch_1500.h5\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x14c9464cb08>"
            ]
          },
          "execution_count": 13,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras.layers import Input, Dense\n",
        "from keras.models import Model\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "\n",
        "# Define the self-supervised model architecture\n",
        "input_layer = Input(shape=(7,))\n",
        "encoder = Dense(128, activation='relu')(input_layer)\n",
        "encoder = Dense(64, activation='relu')(encoder)\n",
        "encoded = Dense(32, activation='relu')(encoder)\n",
        "decoder = Dense(64, activation='relu')(encoded)\n",
        "decoder = Dense(128, activation='relu')(decoder)\n",
        "decoded = Dense(7, activation='linear')(decoder)  # Reconstruction layer\n",
        "\n",
        "# Create the encoder model (used for self-supervised task)\n",
        "encoder_model = Model(inputs=input_layer, outputs=encoded)\n",
        "\n",
        "# Create the full self-supervised model (including reconstruction loss)\n",
        "self_supervised_model = Model(inputs=input_layer, outputs=decoded)\n",
        "\n",
        "# Compile the self-supervised model with a loss function (reconstruction loss)\n",
        "self_supervised_model.compile(optimizer='adam', loss='mean_squared_error')\n",
        "\n",
        "# Train the self-supervised model on the data\n",
        "self_supervised_model.fit(X_train, X_train, epochs=300, batch_size=2048, validation_data=(X_val, X_val), callbacks=[PlotLossesKeras()])\n",
        "\n",
        "# Extract encoded features from the encoder model\n",
        "encoded_features_train = encoder_model.predict(X_train)\n",
        "encoded_features_val = encoder_model.predict(X_val)\n",
        "\n",
        "# Now you can use the encoded features as inputs for a classification head\n",
        "classification_input_layer = Input(shape=(32,))  # Use the appropriate size of encoded features\n",
        "#classification_input_layer_1 =  tf.keras.layers.Dropout(0.5)(classification_input_layer)\n",
        "classification_output_layer = Dense(1, activation='sigmoid')(classification_input_layer)\n",
        "\n",
        "classification_model = Model(inputs=classification_input_layer, outputs=classification_output_layer)\n",
        "classification_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "checkpoint_path = './ssl_ckpt_no_dropout/' + 'weights_epoch_{epoch:02d}.h5'\n",
        "ckpt = tf.keras.callbacks.ModelCheckpoint(\n",
        "        checkpoint_path,\n",
        "        monitor='val_loss',\n",
        "        verbose=1,\n",
        "        save_best_only=False,\n",
        "        save_weights_only=True,\n",
        "        mode='min',\n",
        "        save_freq='epoch'\n",
        "    )\n",
        "\n",
        "# Train the classification model on the encoded features\n",
        "classification_model.fit(encoded_features_train, y_train, epochs=1500, batch_size=2048, validation_data=(encoded_features_val, y_val), callbacks=[PlotLossesKeras(), ckpt])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "100%|██████████| 1500/1500 [01:20<00:00, 18.68it/s]"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Best ROC AUC: 0.95\n",
            "Best accuracy: 0.89\n",
            "Best PR(0): 0.94\n",
            "Best R(0): 0.89\n",
            "Best F1(0): 0.92\n",
            "Best PR(1): 0.77\n",
            "Best R(1): 0.97\n",
            "Best F1(1): 0.82\n",
            "Best F1: 0.89\n",
            "Best precision: 0.89\n",
            "Best recall: 0.89\n",
            "{'auc': './ssl_ckpt_no_dropout/weights_epoch_1481.h5', 'accuracy': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'f1': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'f1_0': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'f1_1': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'precision': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'r0': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'r1': './ssl_ckpt_no_dropout/weights_epoch_01.h5', 'pr_0': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'pr_1': './ssl_ckpt_no_dropout/weights_epoch_1297.h5', 'recall': './ssl_ckpt_no_dropout/weights_epoch_1297.h5'}\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import roc_auc_score, classification_report, accuracy_score, f1_score, precision_score, recall_score\n",
        "from tqdm import tqdm\n",
        "best_auc = 0\n",
        "best_accuracy = 0\n",
        "best_f1 = 0\n",
        "best_f1_0 = 0\n",
        "best_f1_1 = 0\n",
        "best_precision = 0\n",
        "best_pr_0 = 0\n",
        "best_pr_1 = 0\n",
        "best_recall = 0\n",
        "best_r0 = 0\n",
        "best_r1 = 0\n",
        "best_weights = {\n",
        "    'auc': None, \n",
        "    'accuracy': None, \n",
        "    'f1':None, \n",
        "    'f1_0':None, \n",
        "    'f1_1':None,\n",
        "    'precision': None,\n",
        "    'r0': None,\n",
        "    'r1': None,\n",
        "    'pr_0':None,\n",
        "    'pr_1':None,\n",
        "    'recall': None\n",
        "}\n",
        "encoded_features_test = encoder_model.predict(X_test, verbose=0)\n",
        "for epoch in tqdm(range(len(os.listdir('./ssl_ckpt_no_dropout/')))):\n",
        "    weights_path = f\"./ssl_ckpt_no_dropout/weights_epoch_{epoch+1:02d}.h5\"\n",
        "    classification_model.load_weights(weights_path)\n",
        "    \n",
        "    # Evaluate the model\n",
        "    predictions = classification_model.predict(encoded_features_test,verbose=0)\n",
        "    roc_auc = roc_auc_score(y_test, predictions)\n",
        "    accuracy = accuracy_score(y_test, np.round(predictions))\n",
        "    f1_0 = f1_score(y_test, np.round(predictions), pos_label=0)\n",
        "    f1_1 = f1_score(y_test, np.round(predictions), pos_label=1)\n",
        "    f1 = f1_score(y_test, np.round(predictions), average='weighted')\n",
        "    precision = precision_score(y_test, np.round(predictions), average='weighted')\n",
        "    pr_0 = precision_score(y_test, np.round(predictions), pos_label=0)\n",
        "    pr_1 = precision_score(y_test, np.round(predictions), pos_label=1)\n",
        "    recall = recall_score(y_test, np.round(predictions), average='weighted')\n",
        "    r0 = recall_score(y_test, np.round(predictions), pos_label=0)\n",
        "    r1 = recall_score(y_test, np.round(predictions), pos_label=1)\n",
        "    \n",
        "    #accuracy = standard_nn_classifier.evaluate(X_test, y_test, verbose=0)[1]\n",
        "    \n",
        "    if roc_auc > best_auc:\n",
        "        best_auc = roc_auc\n",
        "        best_weights['auc'] = weights_path\n",
        "    if accuracy > best_accuracy:\n",
        "        best_accuracy = accuracy\n",
        "        best_weights['accuracy'] = weights_path\n",
        "    if f1 > best_f1:\n",
        "        best_f1 = f1\n",
        "        best_weights['f1'] = weights_path\n",
        "    if f1_0 > best_f1_0:\n",
        "        best_f1_0 = f1_0\n",
        "        best_weights['f1_0'] = weights_path\n",
        "    if f1_1 > best_f1_1:\n",
        "        best_f1_1 = f1_1\n",
        "        best_weights['f1_1'] = weights_path\n",
        "    if precision > best_precision:\n",
        "        best_precision = precision\n",
        "        best_weights['precision'] = weights_path\n",
        "    if pr_0 > best_pr_0:\n",
        "        best_pr_0 = pr_0 \n",
        "        best_weights['pr_0'] = weights_path\n",
        "    if pr_1 > best_pr_1:\n",
        "        best_pr_1 = pr_1 \n",
        "        best_weights['pr_1'] = weights_path\n",
        "    if r0 > best_r0:\n",
        "        best_r0 = r0 \n",
        "        best_weights['r0'] = weights_path\n",
        "    if r1 > best_r1:\n",
        "        best_r1 = r1 \n",
        "        best_weights['r1'] = weights_path\n",
        "    if recall > best_recall:\n",
        "        best_recall = recall\n",
        "        best_weights['recall'] = weights_path\n",
        "\n",
        "print(\"Best ROC AUC:\", round(best_auc,2))\n",
        "print(\"Best accuracy:\", round(best_accuracy,2))\n",
        "print(\"Best PR(0):\", round(best_pr_0,2))\n",
        "print(\"Best R(0):\", round(best_r0,2))\n",
        "print(\"Best F1(0):\", round(best_f1_0,2))\n",
        "print(\"Best PR(1):\", round(best_pr_1,2))\n",
        "print(\"Best R(1):\", round(best_r1,2))\n",
        "print(\"Best F1(1):\", round(best_f1_1,2))\n",
        "print(\"Best F1:\", round(best_f1,2))\n",
        "print('Best precision:', round(best_precision,2))\n",
        "print('Best recall:', round(best_recall,2))\n",
        "print(best_weights)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0.9536882310814037\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0      0.941     0.894     0.917       142\n",
            "           1      0.773     0.864     0.816        59\n",
            "\n",
            "    accuracy                          0.886       201\n",
            "   macro avg      0.857     0.879     0.866       201\n",
            "weighted avg      0.891     0.886     0.887       201\n",
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import roc_auc_score, classification_report\n",
        "classification_model.load_weights('./ssl_ckpt_no_dropout/weights_epoch_915.h5')\n",
        "predictions = classification_model.predict(encoded_features_test)\n",
        "print(roc_auc_score(y_test, predictions))\n",
        "print(classification_report(y_test, np.round(predictions), digits = 3))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sns.set(style=\"dark\")\n",
        "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
        "cm = confusion_matrix(y_test, np.round(predictions))\n",
        "disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=['BLL', 'FSRQ'])\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0.9511816662687992\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0      0.941     0.894     0.917       142\n",
            "           1      0.773     0.864     0.816        59\n",
            "\n",
            "    accuracy                          0.886       201\n",
            "   macro avg      0.857     0.879     0.866       201\n",
            "weighted avg      0.891     0.886     0.887       201\n",
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import roc_auc_score, classification_report\n",
        "classification_model.load_weights('./ssl_ckpt_no_dropout/weights_epoch_1297.h5')\n",
        "predictions = classification_model.predict(encoded_features_test)\n",
        "print(roc_auc_score(y_test, predictions))\n",
        "print(classification_report(y_test, np.round(predictions), digits = 3))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sns.set(style=\"dark\")\n",
        "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
        "cm = confusion_matrix(y_test, np.round(predictions))\n",
        "disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=['BLL', 'FSRQ'])\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1mxtxTgLAW5a"
      },
      "source": [
        "## Contrastive Classification\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "e8-5iH7EzkUZ",
        "outputId": "6d41ee0a-89e5-4a0f-945e-ac0b7a1c3925"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1200x800 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "accuracy\n",
            "\ttraining         \t (min:    0.505, max:    0.628, cur:    0.627)\n",
            "\tvalidation       \t (min:    0.522, max:    0.652, cur:    0.627)\n",
            "Loss\n",
            "\ttraining         \t (min:    0.659, max:    0.696, cur:    0.659)\n",
            "\tvalidation       \t (min:    0.649, max:    0.690, cur:    0.649)\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x1c0cfe18fc8>"
            ]
          },
          "execution_count": 13,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras.layers import Input, Dense, Concatenate\n",
        "from keras.models import Model\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "# Create pairs of data samples\n",
        "def create_pairs(data, labels):\n",
        "    pairs, pair_labels = [], []\n",
        "    num_samples = data.shape[0]\n",
        "    for i in range(num_samples):\n",
        "        positive_idx = np.where(labels == labels[i])[0]\n",
        "        negative_idx = np.where(labels != labels[i])[0]\n",
        "\n",
        "        positive_sample = data[np.random.choice(positive_idx)]\n",
        "        negative_sample = data[np.random.choice(negative_idx)]\n",
        "\n",
        "        pairs.append([data[i], positive_sample])\n",
        "        pair_labels.append(1)  # Positive pair\n",
        "\n",
        "        pairs.append([data[i], negative_sample])\n",
        "        pair_labels.append(0)  # Negative pair\n",
        "\n",
        "    return np.array(pairs), np.array(pair_labels)\n",
        "\n",
        "pairs_train, pair_labels_train = create_pairs(X_train, y_train)\n",
        "pairs_val, pair_labels_val = create_pairs(X_val, y_val)\n",
        "\n",
        "# Define the contrastive classification model architecture\n",
        "input_a = Input(shape=(7,))\n",
        "input_b = Input(shape=(7,))\n",
        "shared_dense = Dense(64, activation='relu')\n",
        "\n",
        "\n",
        "encoded_a = shared_dense(input_a)\n",
        "encoded_b = shared_dense(input_b)\n",
        "\n",
        "merged = Concatenate()([encoded_a, encoded_b])\n",
        "output_layer = Dense(1, activation='sigmoid')(merged)\n",
        "\n",
        "contrastive_model = Model(inputs=[input_a, input_b], outputs=output_layer)\n",
        "\n",
        "# Compile the contrastive classification model\n",
        "contrastive_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# Train the contrastive classification model on the pairs\n",
        "contrastive_model.fit([pairs_train[:, 0], pairs_train[:, 1]], pair_labels_train,\n",
        "                      epochs=100, batch_size=2048,\n",
        "                      validation_data=([pairs_val[:, 0], pairs_val[:, 1]], pair_labels_val),\n",
        "                      callbacks=[PlotLossesKeras()]\n",
        "                      )\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yCUIo18H1G6o",
        "outputId": "3c8e5562-c4f6-4c97-afc2-64e705eb31ec"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1200x800 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "accuracy\n",
            "\ttraining         \t (min:    0.592, max:    0.913, cur:    0.912)\n",
            "\tvalidation       \t (min:    0.615, max:    0.900, cur:    0.895)\n",
            "Loss\n",
            "\ttraining         \t (min:    0.233, max:    0.769, cur:    0.233)\n",
            "\tvalidation       \t (min:    0.245, max:    0.718, cur:    0.245)\n",
            "\n",
            "Epoch 01500: saving model to ./ssl_contrastive_ckpt\\weights_epoch_1500.h5\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x1c036da0dc8>"
            ]
          },
          "execution_count": 14,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Extract features from the contrastive classification model\n",
        "feature_extraction_model = Model(inputs=contrastive_model.input[0], outputs=encoded_a)  # Use one branch as feature extractor\n",
        "feature_extraction_model.save_weights('./ssl_contrastive.h5')\n",
        "features_train = feature_extraction_model.predict(X_train, verbose = 0)\n",
        "features_val = feature_extraction_model.predict(X_val, verbose = 0)\n",
        "\n",
        "# Build a classification model\n",
        "classification_input_layer = Input(shape=(64,))  # Adjust the input shape to match the size of the extracted features\n",
        "classification_output_layer = Dense(1, activation='sigmoid')(classification_input_layer)\n",
        "\n",
        "classification_model = Model(inputs=classification_input_layer, outputs=classification_output_layer)\n",
        "classification_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "checkpoint_path = './ssl_contrastive_ckpt/' + 'weights_epoch_{epoch:02d}.h5'\n",
        "ckpt = tf.keras.callbacks.ModelCheckpoint(\n",
        "        checkpoint_path,\n",
        "        monitor='val_loss',\n",
        "        verbose=1,\n",
        "        save_best_only=False,\n",
        "        save_weights_only=True,\n",
        "        mode='min',\n",
        "        save_freq='epoch'\n",
        "    )\n",
        "\n",
        "# Train the classification model on your labeled data\n",
        "classification_model.fit(\n",
        "    features_train, \n",
        "    y_train, \n",
        "    epochs=1500, \n",
        "    batch_size=2048, \n",
        "    validation_data=(features_val, y_val),\n",
        "    callbacks=[PlotLossesKeras(), ckpt])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 0/1500 [00:00<?, ?it/s]"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "100%|██████████| 1500/1500 [01:26<00:00, 17.27it/s]"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Best ROC AUC: 0.95\n",
            "Best accuracy: 0.89\n",
            "Best PR(0): 0.94\n",
            "Best R(0): 0.99\n",
            "Best F1(0): 0.92\n",
            "Best PR(1): 0.91\n",
            "Best R(1): 0.86\n",
            "Best F1(1): 0.82\n",
            "Best F1: 0.89\n",
            "Best precision: 0.89\n",
            "Best recall: 0.89\n",
            "{'auc': './ssl_contrastive_ckpt/weights_epoch_1378.h5', 'accuracy': './ssl_contrastive_ckpt/weights_epoch_890.h5', 'f1': './ssl_contrastive_ckpt/weights_epoch_890.h5', 'f1_0': './ssl_contrastive_ckpt/weights_epoch_890.h5', 'f1_1': './ssl_contrastive_ckpt/weights_epoch_890.h5', 'precision': './ssl_contrastive_ckpt/weights_epoch_890.h5', 'r0': './ssl_contrastive_ckpt/weights_epoch_25.h5', 'r1': './ssl_contrastive_ckpt/weights_epoch_225.h5', 'pr_0': './ssl_contrastive_ckpt/weights_epoch_890.h5', 'pr_1': './ssl_contrastive_ckpt/weights_epoch_99.h5', 'recall': './ssl_contrastive_ckpt/weights_epoch_890.h5'}\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import roc_auc_score, classification_report, accuracy_score, f1_score, precision_score, recall_score\n",
        "from tqdm import tqdm\n",
        "best_auc = 0\n",
        "best_accuracy = 0\n",
        "best_f1 = 0\n",
        "best_f1_0 = 0\n",
        "best_f1_1 = 0\n",
        "best_precision = 0\n",
        "best_pr_0 = 0\n",
        "best_pr_1 = 0\n",
        "best_recall = 0\n",
        "best_r0 = 0\n",
        "best_r1 = 0\n",
        "best_weights = {\n",
        "    'auc': None, \n",
        "    'accuracy': None, \n",
        "    'f1':None, \n",
        "    'f1_0':None, \n",
        "    'f1_1':None,\n",
        "    'precision': None,\n",
        "    'r0': None,\n",
        "    'r1': None,\n",
        "    'pr_0':None,\n",
        "    'pr_1':None,\n",
        "    'recall': None\n",
        "}\n",
        "encoded_features_test = feature_extraction_model.predict(X_test, verbose=0)\n",
        "for epoch in tqdm(range(len(os.listdir('./ssl_contrastive_ckpt/')))):\n",
        "    weights_path = f\"./ssl_contrastive_ckpt/weights_epoch_{epoch+1:02d}.h5\"\n",
        "    classification_model.load_weights(weights_path)\n",
        "    \n",
        "    # Evaluate the model\n",
        "    predictions = classification_model.predict(encoded_features_test,verbose=0)\n",
        "    roc_auc = roc_auc_score(y_test, predictions)\n",
        "    accuracy = accuracy_score(y_test, np.round(predictions))\n",
        "    f1_0 = f1_score(y_test, np.round(predictions), pos_label=0)\n",
        "    f1_1 = f1_score(y_test, np.round(predictions), pos_label=1)\n",
        "    f1 = f1_score(y_test, np.round(predictions), average='weighted')\n",
        "    precision = precision_score(y_test, np.round(predictions), average='weighted')\n",
        "    pr_0 = precision_score(y_test, np.round(predictions), pos_label=0)\n",
        "    pr_1 = precision_score(y_test, np.round(predictions), pos_label=1)\n",
        "    recall = recall_score(y_test, np.round(predictions), average='weighted')\n",
        "    r0 = recall_score(y_test, np.round(predictions), pos_label=0)\n",
        "    r1 = recall_score(y_test, np.round(predictions), pos_label=1)\n",
        "    \n",
        "    #accuracy = standard_nn_classifier.evaluate(X_test, y_test, verbose=0)[1]\n",
        "    \n",
        "    if roc_auc > best_auc:\n",
        "        best_auc = roc_auc\n",
        "        best_weights['auc'] = weights_path\n",
        "    if accuracy > best_accuracy:\n",
        "        best_accuracy = accuracy\n",
        "        best_weights['accuracy'] = weights_path\n",
        "    if f1 > best_f1:\n",
        "        best_f1 = f1\n",
        "        best_weights['f1'] = weights_path\n",
        "    if f1_0 > best_f1_0:\n",
        "        best_f1_0 = f1_0\n",
        "        best_weights['f1_0'] = weights_path\n",
        "    if f1_1 > best_f1_1:\n",
        "        best_f1_1 = f1_1\n",
        "        best_weights['f1_1'] = weights_path\n",
        "    if precision > best_precision:\n",
        "        best_precision = precision\n",
        "        best_weights['precision'] = weights_path\n",
        "    if pr_0 > best_pr_0:\n",
        "        best_pr_0 = pr_0 \n",
        "        best_weights['pr_0'] = weights_path\n",
        "    if pr_1 > best_pr_1:\n",
        "        best_pr_1 = pr_1 \n",
        "        best_weights['pr_1'] = weights_path\n",
        "    if r0 > best_r0:\n",
        "        best_r0 = r0 \n",
        "        best_weights['r0'] = weights_path\n",
        "    if r1 > best_r1:\n",
        "        best_r1 = r1 \n",
        "        best_weights['r1'] = weights_path\n",
        "    if recall > best_recall:\n",
        "        best_recall = recall\n",
        "        best_weights['recall'] = weights_path\n",
        "\n",
        "print(\"Best ROC AUC:\", round(best_auc,2))\n",
        "print(\"Best accuracy:\", round(best_accuracy,2))\n",
        "print(\"Best PR(0):\", round(best_pr_0,2))\n",
        "print(\"Best R(0):\", round(best_r0,2))\n",
        "print(\"Best F1(0):\", round(best_f1_0,2))\n",
        "print(\"Best PR(1):\", round(best_pr_1,2))\n",
        "print(\"Best R(1):\", round(best_r1,2))\n",
        "print(\"Best F1(1):\", round(best_f1_1,2))\n",
        "print(\"Best F1:\", round(best_f1,2))\n",
        "print('Best precision:', round(best_precision,2))\n",
        "print('Best recall:', round(best_recall,2))\n",
        "print(best_weights)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0.9513010264979709\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0      0.941     0.894     0.917       142\n",
            "           1      0.773     0.864     0.816        59\n",
            "\n",
            "    accuracy                          0.886       201\n",
            "   macro avg      0.857     0.879     0.866       201\n",
            "weighted avg      0.891     0.886     0.887       201\n",
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import roc_auc_score, classification_report\n",
        "classification_model.load_weights('./ssl_contrastive_ckpt/weights_epoch_890.h5')\n",
        "predictions = classification_model.predict(encoded_features_test)\n",
        "print(roc_auc_score(y_test, predictions))\n",
        "print(classification_report(y_test, np.round(predictions), digits = 3))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sns.set(style=\"dark\")\n",
        "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
        "cm = confusion_matrix(y_test, np.round(predictions))\n",
        "disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=['BLL', 'FSRQ'])\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.15"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}