danheck/gpt: Generalized Processing Tree Models

Fits generalized processing tree models (GPT; Heck, Erdfelder, & Kieslich, in press), a model class that accounts for multivariate data with one discrete and one or more continuous variables. GPT models extend multinomial processing tree (MPT) models that are frequenctly used in psychology by assuming that the distribution of continuous variables is a (multivariate) mixture, in which the mixture probabilities are constrained by the MPT structure of conditional probabilities. The package implements parameter estimation by expectation-maximization (EM) and gradient-based optimization, model tests, summaries, and plots.

install.packages("remotes")
remotes::install_github("danheck/gpt")