AI4SDB
Home/Tools/Research Data Analysis/PyMC

Research Data Analysis

PyMC

An open-source probabilistic-programming library for specifying Bayesian models in Python and fitting them with MCMC or variational inference.

Last verified

RESEARCH USE

Where PyMC fits

PyMC lets researchers express priors, likelihoods, and hierarchical structure in Python, then estimate a posterior with NUTS and other MCMC methods or variational inference. A defensible analysis includes prior predictive checks, chain diagnostics, effective sample sizes, and posterior predictive checks. Preserve the model code, processed data, seeds, and sampling configuration.

Research tasks

  • Specify hierarchical and generalized Bayesian models
  • Estimate posteriors with MCMC or variational methods
  • Run prior and posterior predictive checks

What to evaluate before use

  • Completed sampling is not evidence of a sound model; divergences, poor mixing, non-identifiability, and weak priors can invalidate conclusions.
  • Variational inference is approximate. Its speed does not remove the need to assess approximation quality and posterior coverage.

Verification note

This entry summarizes the tool's role without assessing scientific accuracy or endorsing its outputs. Features and terms can change; consult the official source before adopting it for consequential work.

Bayesian modelingprobabilistic programming

Last verified: 2026-09-05
Source: official documentation ↗