cmdstanpy – Python interface to CmdStan

• Overview
• Installation
  • Install package CmdStanPy
  • Install CmdStan
    • Prerequisites
    • Function install_cmdstan
    • DIY Installation
    • Post Installation: Setting Environment Variables
• “Hello, World”
  • Bayesian estimation via Stan’s HMC-NUTS sampler
    • Instantiate the Stan model, assemble the data
    • Run the HMC-NUTS sampler
    • Access the sample
    • Summarize the results
    • Save the Stan CSV files
• Stan Models in CmdStanPy
  • Model compilation
    • Specifying a custom Make tool
• MCMC Sampling
  • NUTS-HMC sampler configuration
  • CmdStanMCMC: HMC sample and metadata
  • Example: fit model - sampler defaults
  • Example: high-level parallelization with reduce_sum
  • Example: generate data - fixed_param=True
• Run Generated Quantities
  • Configuration
  • Example: add posterior predictive checks to bernoulli.stan
• Maximum Likelihood Estimation
  • Optimize configuration
  • Example: estamate MLE for model bernoulli.stan by optimization
  • References
• Variational Inference
  • ADVI configuration
  • Example: variational inference for model bernoulli.stan
  • References
• Under the Hood
  • File Handling
    • Input Data
    • Output Files
• API Reference
  • Classes
    • CmdStanModel
    • CmdStanMCMC
    • CmdStanMLE
    • CmdStanGQ
    • CmdStanVB
    • RunSet
  • Functions
    • cmdstan_path
    • install_cmstan
    • set_cmdstan_path
    • set_make_env

Index