pymc-modeling

Modern Bayesian modeling with PyMC v5+. Key defaults: nutpie sampler (2-5x faster), non-centered parameterization for hierarchical models, HSGP over exact GPs, coords/dims for readable InferenceData, and save-early workflow to prevent data loss from late crashes.

Modeling strategy: Build models iteratively — start simple, check prior predictions, fit and diagnose, check posterior predictions, expand one piece at a time. See references/workflow.md for the full workflow.

Notebook preference: Use marimo for interactive modeling unless the project already uses Jupyter.

Modelado estadístico bayesiano con PyMC v5+. Úselo al crear modelos probabilísticos, especificar antecedentes, ejecutar inferencia MCMC, diagnosticar convergencia o comparar modelos. Cubre los backends PyMC, ArviZ, pymc-bart, pymc-extras, nutpie y JAX/NumPyro. Activadores de tareas que involucran: inferencia bayesiana, muestreo posterior, modelos jerárquicos/multinivel, GLM, series temporales, procesos gaussianos, BART, modelos mixtos, verificaciones predictivas previas/posteriores, diagnósticos MCMC, LOO-CV, WAIC, comparación de modelos o inferencia causal con hacer/observar. Fuente: pymc-labs/python-analytics-skills.