mixturemodelsr - R Wrapper for the Mixture-Models Python Library

First R package enabling mixture models for high-dimensional data through gradient-based optimization with Automatic Differentiation (AD). Provides an R interface to the 'Mixture-Models' Python package (Kasa et al., 2024) via reticulate. Unlike traditional EM-based approaches (e.g., mclust, flexmix), this package uses AD and gradient-based optimization (including second-order Newton-CG) to fit Gaussian Mixture Models (GMM), Mixture of Factor Analyzers (MFA), Parsimonious GMM (PGMM), MCLUST family, and t-mixture models without requiring stringent modeling constraints, making it suitable for high-dimensional settings where the number of parameters exceeds the sample size. Reference: Kasa, S. R., Yijie, H., Kasa, S. K., & Rajan, V. (2024). Mixture-Models: a one-stop Python Library for Model-based Clustering using various Mixture Models. arXiv preprint arXiv:2402.10229.