xifa 0.1.1

Accelerated Item Factor Analysis

XIFA: Accelerated Item Factor Analysis

xifa is a python package for conducting item factor analysis (IFA), a representative multivariate techniques in psychometrics.

xifa is build on jax, a package for Autograd and XLA. Hence, xifa can run IFA on GPUs and TPUs to speed up the training process. That is why we call it Accelerated IFA.

xifa implements a vectorized Metropolis-Hastings Robbins-Monro (MH-RM) algorithm to calculate a marginal maximum likelihood (MML) estimate. MH-RM is one of the states-of-art algorithms for high dimensional IFA (Cai, 2010). The vectorized version is designed for parallel computing with GPUs of TPUs. The vectorized MH-RM includes two stages. The first stage updates the parameter estimate by a stochastic expectation-maximization (StEM) algorithm. The second stage conducts stochastic approximation (SA) to refine the estimate.

In the current version (0.1.1), xifa supports ordinal data IFA with graded response model (GRM; Semejima, 1969) and generalized partial credit model (GPCM; Muraki, 1992). The analysis can be either exploratory or confirmatory.

For a tutorial, please see IPIP 50 Items Example.