Portable Models and Research Artifacts
In gp3ml: Governance-First Predictive Modelling for 'Gazepoint' Research

knitr::opts_chunk$set(collapse = TRUE, comment = "#>")
library(gp3ml)

A reproducible model is more than an R object. This layer records the model, preprocessing, task, feature provenance, decision rule, schema, environment, and cryptographic file hashes. Optional bundle support is used when available for engines that need custom serialization.

Environment records can be compared before reproducing a result:

reference <- capture_gazepoint_environment(packages = "gp3ml")
comparison <- compare_gazepoint_environments(reference, reference)
comparison$status
plot(comparison)

A minimal RO-Crate-oriented export can package research files with SHA-256 hashes. gp3ml deliberately describes this as RO-Crate-oriented unless an independent conformance validator has also been run.