Package: psychnets 0.2.1

psychnets: Clean-Room Base-R Psychometric Network Estimation

Estimates psychometric network models -- correlation and partial correlation, the EBIC-regularized Gaussian graphical model (graphical lasso), its nonparanormal and unregularized stepwise variants, information-filtering graphs (TMFG and LoGo), relative-importance networks, and the Ising and mixed graphical models -- reimplemented from first principles in base R with no compiled dependencies. Each regularized estimator ships a dependency-free correctness certificate (for the Gaussian graphical model, the graphical-lasso stationarity / KKT residual) so a fitted network is self-verifying: its distance from the unique optimum of its own convex objective is reported directly, rather than trusted only because it matches an external solver. The R counterpart to the 'psychaj' TypeScript library.

Authors:Mohammed Saqr [aut, cre]