envdataprep 0.1.0

Extensible Environmental Data Preprocessing Framework

EnvDataPrep: High-performance Environmental Data Pre-processing

Why EnvDataPrep?

EnvDataPrep saves Money, Time and Disk Storage for those who deal with environmental datasets.

Current Capacity

The first feature is the Unified, Configuration-Driven Extraction of NetCDF Files. Below is an example usage that reduces a TROPOMI satellite product file size by ~90%:

"""Example of extracting a subset of variables from a netCDF file."""

from pathlib import Path
import envdataprep as edp

# Set up input and output directories
ROOT = Path("E:/Samples/Satellites")
input_dir = ROOT / "Input"
output_dir = ROOT / "Output"

# Select an example file
file_name = "S5P_RPRO_L2__NO2____20190101T233659_20190102T011828_06322_03_020400_20221106T093236.nc"
file_path = input_dir / file_name

# List all variables in the file
variables = edp.list_netcdf_variables(file_path)
print(*variables, sep='\n')

# List the variables to extract
variable_paths = [
    "PRODUCT/latitude",
    "PRODUCT/longitude",
    "PRODUCT/nitrogendioxide_tropospheric_column",
    "PRODUCT/SUPPORT_DATA/INPUT_DATA/cloud_fraction_crb",
]

# Extract and save out
# By default, the output file preserves the original group structure
edp.subset_netcdf(
    file_path,
    output_dir,
    variable_paths,
    output_name="example_extracted_data.nc",
    compression_method='zlib',
    compression_level=9,
)

Installation

Prerequisites

• Mamba (recommended) or Conda
  • Preferred for installing scientific Python dependencies (netCDF4, xarray, numpy)
  • Handles complex dependency resolution more reliably than pip alone
• Alternative: pip installation should also work, but will be more complicated

Quick Setup

# 1. Get the code
git clone https://github.com/envmini/envdataprep.git
cd envdataprep

# 2. Create environment (choose one)
# Option A: Using Mamba (faster, recommended)
mamba env create -f environment.yml
mamba activate envdataprep

# Option B: Using Conda
# conda env create -f environment.yml
# conda activate envdataprep

# 3. Install package in development mode
pip install -e . --no-deps

# 4. Verify installation
python -c "import envdataprep; print('Installation successful!')"

Why development mode for now?

• Package not yet published to PyPI/conda-forge
• Allows you to get latest features and contribute feedback
• Easy to update with git pull

Note: We use pip install -e . even in conda/mamba environments, but this pip command uses the pip from your active conda environment, not system pip, You can verify this with:

which pip  # Should show: .../miniforge3/envs/envdataprep/bin/pip

License

This project is licensed under the MIT License.

⬆ Back to top