I have contributed to three publically available statistical software packages, and I hope to add more in the future.
momentLS
Autocovariance estimation for reversible Markov chains. The package implements the estimators in “Efficient shape-constrained inference for the autocovariance sequence from a reversible Markov chain” ( https://arxiv.org/abs/2207.12705 ). I am a contributor to this package, with Hyebin Song.
automultinomial
For regression models (similar to logistic regression) of spatially correlated discrete data.
bcd
This is a C++ implementation (with R wrappers via Rcpp) of block coordinate descent for fitting group lasso penalized linear (logistic, multinomial, Poisson, multiresponse linear) regression models. It supports the overlapping group lasso without explicit duplication of the columns of the design matrix, and sparse design matrices.