External IRT engines and exact-method gating

In eyeprocess: Harmonize Eye-Tracking, Pupillometry, Biometrics, and Psychometric Process Data

knitr::opts_chunk$set(collapse = TRUE, comment = "#>")
library(eyeprocess)

The engine registry makes estimation boundaries inspectable.

eyeprocess_irt_engine_registry()

The current registry covers mirt, TAM, GDINA, LNIRT, eRm, equateIRT, catR, and mirtCAT. Wrappers return an eye_external_irt_fit when the requested package is available; otherwise they return an eye_gated_irt_engine with fit = NULL. They never switch to another estimator.

fit <- fit_eyeprocess_mirt(response_matrix, model = 1, itemtype = "2PL")
validate_eyeprocess_external_irt_fit(fit, engine = "mirt")

This is the same governance principle used for eyeprocess's existing frontier estimators: estimator identity is part of the scientific specification.

eyeprocess documentation built on Sept. 28, 2026, 5:08 p.m.