NonnegMFPy solves nonnegative matrix factorization (NMF) given a dataset with heteroscedastic uncertainties and missing data with a vectorized multiplicative update rule; this can be used create a mask and iterate the process to exclude certain new data by updating the mask. The code can work on multi-dimensional data, such as images, if the data are first flattened to 1D.

SetCoverPy finds an (near-)optimal solution to the set cover problem (SCP) as fast as possible. It employs an iterative heuristic approximation method, combining the greedy and Lagrangian relaxation algorithms. It also includes a few useful tools for a quick chi-squared fitting given two vectors with measurement errors.

redmonster performs automated redshift measurement, physical‑parameter estimation, and spectral classification of 1D astronomical spectra. This set of Python utilities outputs the best‑fit model, redshift, classification, and derived parameters in standard formats for downstream analysis. The repository includes templates, configuration files, and documentation to enable flexible redshift and parameter estimation workflows.