Scaffold Bayesian workflows with NumPyro and ArviZ diagnostics.
npx skills add https://github.com/lowmason/agent-skills --skill bayesian-workflow
Please help me install this Agent Skill. Skill: bayesian-workflow Source: https://github.com/lowmason/agent-skills/tree/main/bayesian-workflow Command: npx skills add https://github.com/lowmason/agent-skills --skill bayesian-workflow
Guides and guards your Bayesian workflow from data to decision, providing an end-to-end scaffold for building, diagnosing, and reporting probabilistic models with NumPyro (JAX) and ArviZ.
Install the bayesian-workflow skill by placing the bayesian-workflow folder in your skills directory and running the main.py to initialize the Bayesian workflow.
High-intent search queries and answers about installing and using this skill.
An end-to-end Bayesian workflow with NumPyro and ArviZ involves specifying priors, running inference, performing convergence diagnostics like R-hat and ESS, and completing posterior predictive checks before saving InferenceData to disk.
When building Bayesian models with NumPyro, you should run prior predictive checks, convergence diagnostics including R-hat and ESS, posterior predictive checks, LOO-CV, and calibration plots to ensure model validity and reproducibility.
To ensure reproducibility when saving Bayesian inference results, you should save the generated InferenceData object to disk after sampling and use programmatic reporting templates to produce canonical report artifacts.
Yes, you can use ArviZ to perform LOO-CV and posterior predictive checks on NumPyro models, applying cross-tool compatibility to evaluate model calibration and generate diagnostic plots within the workflow.
The best way to report Bayesian posterior results programmatically is to use canonical reporting templates that generate structured report artifacts directly from your saved InferenceData after sampling.
Yes, you need JAX installed because NumPyro is built on top of JAX to enable probabilistic inference, and the workflow relies on ArviZ and NumPy dependencies to execute diagnostics and reporting.
ovachieverBuild, fit, and evaluate Bayesian models with PyMC using MCMC sampling.
cartaExport Carta report data to branded Excel files with live aggregation formulas.
xjtulycBuild Bayesian models with PyMC 5.x and ArviZ using NUTS sampling.
galeepBuild and analyze Bayesian models with PyMC for MCMC and variational inference.
