PyNGauss is a Python package for computing non-Gaussian summary statistics of 2D random fields and sky maps. Its long-term goal is to provide a unified framework for morphology-based and harmonic-space estimators, including Minkowski Functionals, skewness–kurtosis statistics, bispectra, and related tools. The current release includes implementations of Minkowski Functionals and generalized skewness–kurtosis statistics.
You can install PyNGauss directly from the source code. First, clone the repository to your local machine:
git clone [https://github.com/fazlurahman/pyngauss.git](https://github.com/fazlurahman/pyngauss.git)
cd pyngaussTo install the package normally, run:
pip install .If you plan to modify the code or run notebooks directly from the repository, install it in editable mode so your changes take effect immediately:
pip install -e .If you use PyNGauss in your research, please cite the primary package paper:
- Rahman, et al. (2026). Understanding the non-Gaussian nature of Galactic foreground emissions towards small scales
The theoretical framework and numerical algorithms implemented in this code are based on the following foundational papers:
- Continuous Sphere Theory: Schmalzing, J., & Górski, K. M. (1998). Minkowski Functionals used in the Morphological Analysis of Cosmic Microwave Background Anisotropy Maps Monthly Notices of the Royal Astronomical Society, 297(2), 355-365. arXiv:astro-ph/9710185
- Pixel Window Corrections (Lim-Simon): Lim, E. A., & Simon, D. (2012). Can we detect Hot or Cold spots in the CMB with Minkowski Functionals?. Journal of Cosmology and Astroparticle Physics, 2012(01), 048. arXiv:1103.4300