the Grism Emission-line Kinematics tOol

Description

Geko is a Python package for analyzing grism spectroscopy from JWST NIRCam observations. The package uses JAX for accelerated computation and Numpyro for Bayesian inference to recover emission-line kinematics and morphology from 2D slitless spectroscopy spectra.

Key Features:

• JAX-accelerated: GPU support for fast inference
• Bayesian inference: MCMC fitting using Numpyro's No-U-Turn Sampler (NUTS)
• Flexible configuration: Easy-to-use configuration system for priors and MCMC parameters
• Comprehensive visualization: Diagnostic plots and corner plots for fit results

Installation

Using pip (Recommended)

pip install astro-geko

Using Conda

For a complete environment with all dependencies:

# Clone the repository
git clone https://github.com/angelicalola-danhaive/geko.git
cd geko

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

# Install geko
pip install astro-geko

Development Installation

If you want to install the development version:

# Clone the repository
git clone https://github.com/angelicalola-danhaive/geko.git
cd geko

# Install in editable mode
pip install -e .

Requirements

• Python >= 3.8
• JAX/JAXlib (with optional GPU support)
• Numpyro
• Astropy
• Photutils
• PySersic

For GPU acceleration, follow the JAX installation guide for CUDA support.

Quick Start

from geko.fitting import run_geko_fit

# Run a fit with minimal configuration
inference_data = run_geko_fit(
    output='my_fit',
    master_cat='path/to/catalog.cat',
    line='Ha',
    parametric=True,
    save_runs_path='./saves/',
    num_chains=2,
    num_warmup=500,
    num_samples=1000,
    source_id=12345,
    field='GOODS-S-FRESCO',
    grism_filter='F444W'
)

See the documentation for detailed usage examples and tutorials.

Citation

If you use Geko in your research, please cite the following paper:

@article{Danhaive:2025ac,
	author = {{Danhaive}, A. Lola and {Tacchella}, Sandro},
	journal = {arXiv e-prints},
	month = oct,
	pages = {arXiv:2510.07369},
	title = {{Modelling the kinematics and morphology of galaxies in slitless spectroscopy  with _geko_}},
	year = 2025}

Acknowledgements

We acknowledge support from the Royal Society Research Grants (G125142). We thank Amanda Stoffers for creating our beautiful logo.

This package makes use of:

• JAX for accelerated numerical computing
• Numpyro for probabilistic programming
• Astropy for astronomical data handling

License

This project is licensed under the MIT License - see the LICENSE file for details.