Estimation of latent variable models using Bayesian methods. Currently estimates the loglinear cognitive diagnosis model of Henson, Templin, and Willse (2009) .
blatent
Bayesian Latent Variable Models in R
Changelist
Version 0.1.2RC1
Fixed bug in syntax reading that did not allow for multiple non-list variables on left-hand side of model equation
Fixed bug in calculation of latent variable summaries for BayesNets specifications
Implemented check for observed variables that are listed in model but are non-numeric class in input data
Created two plot functions:
plot(model, type="fit-heatmap") for heatmaps of discrepancies when PPMC covariance, Pearson correlation, and/or tetrachoric correlation are used
plot(model, type="nonconverged", PSRFcut = 1.1) for plotting trace and density plots of MCMC chains for model parameters with PSRF > PSRFcut
Fixed an issue where joint variables could have appeared in model statements as DVs
Fixed an issue where DVs could be listed multiple times