Stan Development

关于

This skill streamlines the development of Bayesian models using the Stan probabilistic programming language. It enforces modern syntax standards introduced in Stan 2.26+, promotes modular code organization through file inclusions, and offers best practices for integrating Stan with R (cmdstanr) and Python (cmdstanpy). From optimizing performance via vectorization to debugging common MCMC issues like divergences and slow mixing, it serves as a comprehensive companion for researchers and data scientists building robust statistical models for Bayesian inference.

主要功能

• Integration workflows for cmdstanr, including model compilation and caching
• Debugging strategies for MCMC diagnostics, divergences, and sampling issues
• Modular model organization using functions and file inclusions
• Performance optimization through vectorization and efficient matrix operations
• 5 GitHub stars
• Implementation of modern Stan array syntax and constrained data types

使用场景

• Migrating legacy Stan code to modern, high-performance syntax standards
• Developing and unit testing custom Stan functions within R or Python environments
• Building and scaling Bayesian hierarchical models for statistical analysis