Setting up a local Python environment

When extending Caravan to new basins, you will need to process the output of Google Earth Engine locally and merge it with streamflow data. This is done in Caravan_part2_local_postprocessing.ipynb. To be able to run this notebook locally, you need a Python environment with the following specification:

• Python >= 3.7
• jupyter
• netcdf4
• numba
• pandas
• tqdm
• xarray
• timezonefinder[numba]

To help you get started, we provided two different ways how to prepare your Python environment locally.

Using Anaconda/Miniconda

Our recommended way is to use Anaconda or Miniconda for managing different Python environments locally. If you have either of those installed locally, you can create a Python environment from the Terminal/Bash by the following command (assuming you are within the Caravan root directory):

conda env create -f environments/environment.yml

Afterwards you can activate the environment by:

conda activate caravan

Using default Python or Venv

From within your virtual environment or default Python environment, you can install the required libraries by executing

pip install -r environments/requirements.txt

Starting JuPyter

Assuming that everything above was done correctly, you can start your local JuPyter server by running

jupyter notebook

This will open a tab in your browser from which you open the Caravan_part2_local_postprocessing.ipynb notebook and start working.