Merge pull request #115 from Perfexionists/typing

Add typing annotation to Perun

.github/workflows/actions/setup/action.yml

@@ -0,0 +1,38 @@
+name: Setup Environment
+description: Performs setup of Perun, Python and its dependencies
+inputs:
+  python-version:
+    description: 'Python version to run'
+    required: true
+
+runs:
+  using: 'composite'
+  steps:
+    - name: Set up Python ${{ inputs.python-version }}
+      uses: actions/setup-python@v4
+      with:
+        python-version: ${{ inputs.python-version }}
+        cache: 'pip'
+
+    - name: Install Unix dependencies
+      shell: sh
+      run: |
+        sudo apt-get -qq update
+        sudo apt-get install time libunwind8-dev g++-9 gcc-9 git
+        sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-9 90 --slave /usr/bin/g++ g++ /usr/bin/g++-9 --slave /usr/bin/gcov gcov /usr/bin/gcov-9
+
+    - name: Initialize Git
+      shell: sh
+      run: |
+        git config --global user.email "[email protected]"
+        git config --global user.name "Perun Allmighty"
+
+    - name: Upgrade pip and setuptools dependencies
+      shell: sh
+      run: |
+        python -m pip install --upgrade pip setuptools
+
+    - name: Install Tox
+      shell: sh
+      run: |
+        pip install tox

.github/workflows/ubuntu.yml

@@ -14,34 +14,15 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.8", "3.9", "3.10", "3.11"]
+        python-version: ["3.9", "3.10", "3.11"]
 
     steps:
       - uses: actions/checkout@v3
 
-      - name: Set up Python ${{ matrix.python-version }}
-        uses: actions/setup-python@v4
+      - name: Setup Python, Ubuntu and Python environment
+        uses: ./.github/workflows/actions/setup
         with:
           python-version: ${{ matrix.python-version }}
-          cache: 'pip'
-
-      - name: Install Unix dependencies
-        run: |
-          sudo apt-get install time libunwind8-dev g++-9 gcc-9 git
-          sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-9 90 --slave /usr/bin/g++ g++ /usr/bin/g++-9 --slave /usr/bin/gcov gcov /usr/bin/gcov-9
-
-      - name: Initialize Git
-        run: |
-          git config --global user.email "[email protected]"
-          git config --global user.name "Perun Allmighty"
-
-      - name: Upgrade pip and setuptools dependencies
-        run: |
-          python -m pip install --upgrade pip setuptools
-
-      - name: Install Tox
-        run: |
-          pip install tox
 
       - name: Execute tests for Python ${{ matrix.python-version }} using Tox
         run: tox -e py
@@ -54,4 +35,23 @@ jobs:
         uses: codecov/codecov-action@v3
         with:
           flags: coverage-${{ matrix.python-version }}
-          verbose: true
+          verbose: true
+
+  typing:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.9", "3.10", "3.11"]
+
+    steps:
+      - uses: actions/checkout@v3
+
+      - name: Setup Python, Ubuntu and Python environment
+        uses: ./.github/workflows/actions/setup
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Check type correctness for Python ${{ matrix.python-version }} using Tox
+        run: |
+          tox -e typing

Makefile

@@ -1,5 +1,8 @@
 test:
-	python3 -m pytest --cov=./ --cov-report term-missing:skip-covered tests/
+	python3 -m pytest --durations=10 --cov=./ --cov-report term-missing:skip-covered tests/
+
+check:
+	mypy perun/
 
 # Setuptools fails for nested requirements file when installed as `pip install .`, so sadly no
 # simple "dev" optional dependency

README.md

@@ -52,6 +52,9 @@ makes it a good store of profiles along with the context.
 Installation
 ------------
 
+Note that we are no longer maintaining support for Python 3.8. Perun may work, but we strongly 
+advise to upgrade your Python to newer version.
+
 You can install Perun as follows:
 
     git clone https://github.com/tfiedor/perun.git

codecov.yml

@@ -1,4 +1,5 @@
 ignore:
   - "perun/collect/trace"
+  - "perun/thirdparty"
   - "tests/test_tracer"
 

perun/__init__.py

@@ -13,6 +13,8 @@
 
 Perun currently exists as CLI application, with GUI application being in development.
 """
+from __future__ import annotations
+
 import importlib.metadata
 
 __version__ = importlib.metadata.version(__package__)

perun/check/average_amount_threshold.py

@@ -34,19 +34,22 @@
 changed from about six seconds to hundred miliseconds. For these detected changes we report no
 confidence at all.
 """
+from __future__ import annotations
+
+from typing import Any, Iterable
 
 import perun.profile.convert as convert
-import perun.check.factory as check
 import perun.postprocess.regression_analysis.tools as tools
 
-from perun.utils.structs import DegradationInfo
+from perun.utils.structs import DegradationInfo, PerformanceChange
+from perun.profile.factory import Profile
 
 
 DEGRADATION_THRESHOLD = 2.0
 OPTIMIZATION_THRESHOLD = 0.5
 
 
-def get_averages(profile):
+def get_averages(profile: Profile) -> dict[str, float]:
     """Retrieves the averages of all amounts grouped by the uid
 
     :param profiles.Profile profile: dictionary representation of profile
@@ -59,7 +62,9 @@ def get_averages(profile):
     return data_frame.groupby('uid').mean(numeric_only=True).to_dict()['amount']
 
 
-def average_amount_threshold(baseline_profile, target_profile, **_):
+def average_amount_threshold(
+        baseline_profile: Profile, target_profile: Profile, **_: Any
+) -> Iterable[DegradationInfo]:
     """Checks between pair of (baseline, target) profiles, whether the can be degradation detected
 
     This is based on simple heuristic, where for the same function models, we only check the order
@@ -81,11 +86,11 @@ def average_amount_threshold(baseline_profile, target_profile, **_):
         if baseline_average is not None:
             difference_ratio = tools.safe_division(target_average, baseline_average)
             if difference_ratio >= DEGRADATION_THRESHOLD:
-                change = check.PerformanceChange.Degradation
+                change = PerformanceChange.Degradation
             elif 0.0 < difference_ratio <= OPTIMIZATION_THRESHOLD:
-                change = check.PerformanceChange.Optimization
+                change = PerformanceChange.Optimization
             else:
-                change = check.PerformanceChange.NoChange
+                change = PerformanceChange.NoChange
 
             yield DegradationInfo(
                 res=change,

perun/check/best_model_order_equality.py

@@ -39,11 +39,14 @@
 value `1.0` (which would mean, that the model precisely fits the measured values), this signifies
 that the best model fit the data tightly and hence the detected optimization is **not spurious**.
 """
+from __future__ import annotations
+
+from typing import Any, Iterable
 
-import perun.check.factory as check
 import perun.check.general_detection as detection
 
-from perun.utils.structs import DegradationInfo
+from perun.utils.structs import DegradationInfo, PerformanceChange
+from perun.profile.factory import Profile
 
 
 CONFIDENCE_THRESHOLD = 0.9
@@ -57,7 +60,9 @@
 ]
 
 
-def best_model_order_equality(baseline_profile, target_profile, **_):
+def best_model_order_equality(
+        baseline_profile: Profile, target_profile: Profile, **_: Any
+) -> Iterable[DegradationInfo]:
     """Checks between pair of (baseline, target) profiles, whether the can be degradation detected
 
     This is based on simple heuristic, where for the same function models, we only check the order
@@ -80,12 +85,12 @@ def best_model_order_equality(baseline_profile, target_profile, **_):
                 baseline_ordering = MODEL_ORDERING.index(best_baseline_model.type)
                 target_ordering = MODEL_ORDERING.index(best_model.type)
                 if baseline_ordering > target_ordering:
-                    change = check.PerformanceChange.Optimization
+                    change = PerformanceChange.Optimization
                 else:
-                    change = check.PerformanceChange.Degradation
+                    change = PerformanceChange.Degradation
                 degradation_rate = target_ordering - baseline_ordering
             else:
-                change = check.PerformanceChange.NoChange
+                change = PerformanceChange.NoChange
                 degradation_rate = 0
 
             yield DegradationInfo(

perun/check/exclusive_time_outliers.py

@@ -63,24 +63,25 @@
 reported as the IQR multiple of `110.46`.
 
 """
+from __future__ import annotations
 
-from typing import Optional, Generator, Dict, List, Tuple
+from typing import Optional, Iterable, Any
 from difflib import get_close_matches
 
 import pandas as pd
 import numpy as np
 from scipy import stats
 
 from perun.utils.structs import DegradationInfo
-from perun.check.factory import PerformanceChange
 from perun.profile import convert
 from perun.logic import config
 
+from perun.utils.structs import PerformanceChange
 from perun.profile.factory import Profile
 
 
-OldLocMap = Dict[str, str]
-NewLocMap = Dict[str, str]
+OldLocMap = dict[str, str]
+NewLocMap = dict[str, str]
 
 
 MZS_CORRECTION = 0.6745
@@ -91,8 +92,8 @@
 
 
 def exclusive_time_outliers(
-        baseline_profile: Profile, target_profile: Profile, **_
-) -> Generator[DegradationInfo, None, None]:
+        baseline_profile: Profile, target_profile: Profile, **_: Any
+) -> Iterable[DegradationInfo]:
     """Checks the pair of (baseline, target) profiles for changes in function exclusive times.
 
     The method works by detecting 'exclusive time delta' outliers and classifying their severity
@@ -147,14 +148,14 @@ def __init__(self, baseline_profile: Profile, target_profile: Profile) -> None:
         )
         if self.location_filter == "*":
             self.location_filter = None
-        self.cut_off: Optional[float] = float(config.lookup_key_recursively(
+        self.cut_off: float = float(config.lookup_key_recursively(
             "degradation.cutoff", "0.0")
         )
         self.df: pd.DataFrame = self._merge_and_diff(
             self._prepare_profile(baseline_profile), self._prepare_profile(target_profile)
         )
 
-    def detect_changes(self) -> Generator[DegradationInfo, None, None]:
+    def detect_changes(self) -> Iterable[DegradationInfo]:
         """Detect and report the exclusive time changes and the overall degradation / optimization.
 
         The method runs three outlier detection methods, where z-score flags the most significant
@@ -217,7 +218,7 @@ def detect_changes(self) -> Generator[DegradationInfo, None, None]:
             )
 
     @staticmethod
-    def _determine_result_and_confidence(row: pd.Series) -> Tuple[PerformanceChange, str, float]:
+    def _determine_result_and_confidence(row: pd.Series[Any]) -> tuple[PerformanceChange, str, float]:
         """Select the severity, confidence type and confidence rate of the exclusive time change.
 
         :param row: DataFrame row of specific 'uid' (i.e., function) record
@@ -363,7 +364,7 @@ def _merge_and_diff(
 
         :return: the merged and extended DataFrame
         """
-        def _delta_exc(row: pd.Series) -> float:
+        def _delta_exc(row: pd.Series[Any]) -> float:
             """ Helper function for properly computing the absolute exclusive time delta even
             when the function is new / deleted (in those cases, one of the exclusive times is
             nan, which breaks the computation).
@@ -416,8 +417,8 @@ def _delta_exc(row: pd.Series) -> float:
 
 
 def _map_similar_names(
-        strings_old: List[str], strings_new: List[str]
-) -> Tuple[OldLocMap, NewLocMap]:
+        strings_old: list[str], strings_new: list[str]
+) -> tuple[OldLocMap, NewLocMap]:
     """Map profile location names that might have slightly changed in different versions.
 
     E.g., due to the names containing the version number (mylib-3.4 vs mylib-3.5).