Simulate prior distributions in Bayesian factor analysis

Simulate prior distributions in Bayesian factor analysis Explore how to simulate prior distributions in Bayesian factor analysis using the BayesFM R package. In this video, Aaron Smith walks through two key functions—simul.nfac.prior() and simul.R.prior()—to demonstrate how prior assumptions shape the number of latent factors and their correlation structure. Topics covered: - Sampling from the prior distribution of latent factors using Dirichlet-based accept/reject schemes - Identification restrictions and the role of Nid, Kmax, and kappa - Simulating prior correlation matrices via Inverse-Wishart and Huang-Wand priors - Comparing prior specifications across multiple parameter sets - Visualizing acceptance rates and prior distributions Whether you're refining your Bayesian models or exploring prior sensitivity, this walkthrough offers hands-on insights with reproducible R code and clear visualizations. Package used: BayesFM on CRAN Ideal for: statisticians, data scientists, and researchers working with latent variable models and Bayesian inference.