Installation

Contents

Installation#

Requirements#

Geko requires Python >= 3.8 and has the following core dependencies:

JAX/JAXlib: Accelerated numerical computing
Numpyro: Probabilistic programming and MCMC sampling
Astropy: Astronomical data handling
Photutils: Photometry and source detection
PySersic: Sérsic profile fitting

Installation Methods#

Using Conda (Recommended)#

The easiest way to install Geko is using the provided conda environment file:

# Clone the repository
git clone https://github.com/yourusername/geko.git
cd geko

# Create conda environment
conda env create -f environment.yml
conda activate geko_env

# Install geko in development mode
pip install -e .

Development Installation#

For development work, install in editable mode:

# Clone the repository
git clone https://github.com/yourusername/geko.git
cd geko

# Install in editable mode with dev dependencies
pip install -e .
pip install pytest  # For running tests

Verifying Installation#

To verify that Geko is correctly installed:

import geko
from geko.grism import Grism
from geko.fitting import run_geko_fit

print("Geko successfully imported!")

Running Tests#

To run the test suite:

pytest tests/

Optional Dependencies#

GPU Support: For GPU acceleration, install JAX with CUDA support following the JAX installation guide
Visualization: Additional plotting capabilities may require matplotlib and arviz