Skip to content

LOO Influence and Predictive Model Comparison

Python-facing port of loo-influence-and-model-comparison.Rmd from the frozen R gp3bayes 0.5.0 reference. The statistical and governance framing below follows the canonical vignette; executable Python workflows use the mapped APIs listed later.

The LOO display layer converts existing gp3bayes_psis_loo, gp3bayes_loo_comparison, and gp3bayes_loo_weights objects into explicit tables and publication-oriented figures.

For multiple prespecified models:

ELPD differences and predictive weights are retained as descriptive predictive evidence. The package does not promote the highest-ranked model to an automatically preferred substantive model.

Python API mapping

  • gp3bayespy.compare_psis_loo
  • gp3bayespy.compute_loo_model_weights
  • gp3bayespy.compute_psis_loo
  • gp3bayespy.loo_diagnostic_table
  • gp3bayespy.loo_summary_table
  • gp3bayespy.model_comparison_table
  • gp3bayespy.model_weights_table
  • gp3bayespy.plot_loo_influence
  • gp3bayespy.plot_model_comparison
  • gp3bayespy.plot_model_weights

Python usage

import gp3bayespy as gp

# All functions listed above are available from the package root.
# Use help(gp.<function>) or the API reference for the exact Python signature.

An executable workflow for this family is included in ../../examples/predictive_diagnostics.py.