Installation¶

Requirements¶

Python 3.9 or higher
Google Earth Engine account with authenticated access

Install with pip¶

pip install pycropwat

Install from source¶

git clone https://github.com/montimaj/pyCropWat.git
cd pyCropWat
pip install -e .

Install with conda¶

# Create environment from file
conda env create -f environment.yml
conda activate pycropwat

# Install package
pip install -e .

# Or with interactive map support (leafmap, localtileserver)
pip install -e ".[interactive]"

Dependencies¶

The following packages are automatically installed:

Package	Version	Description
earthengine-api	≥0.1.370	Google Earth Engine Python API
numpy	≥1.21.0	Numerical computing
xarray	≥2022.3.0	N-D labeled arrays
rioxarray	≥0.14.0	Rasterio xarray extension
geopandas	≥0.12.0	Geospatial pandas
shapely	≥2.0.0	Geometric operations
dask	≥2022.1.0	Parallel computing
distributed	≥2022.1.0	Dask distributed scheduler
rasterio	≥1.3.0	Raster I/O

Google Earth Engine Authentication¶

Before using pyCropWat, you must authenticate with Google Earth Engine:

earthengine authenticate

This will open a browser window to complete the OAuth flow. For server environments, use:

earthengine authenticate --auth_mode=notebook

Using a GEE Project¶

If you have a specific GEE project, you can specify it:

from pycropwat import EffectivePrecipitation

ep = EffectivePrecipitation(
    asset_id='ECMWF/ERA5_LAND/MONTHLY_AGGR',
    precip_band='total_precipitation_sum',
    geometry_path='study_area.geojson',
    start_year=2020,
    end_year=2023,
    gee_project='my-gee-project'  # Your GEE project ID
)

Or via CLI:

export GOOGLE_CLOUD_PROJECT=my-gee-project
pycropwat --asset ... --gee-geometry ...

Verifying Installation¶

import pycropwat
print(pycropwat.__version__)

# Test GEE connection
import ee
ee.Initialize()
print("GEE connection successful!")