cmdstanpy – Python interface to CmdStan

CmdStanPy is a lightweight interface to Stan for Python users which provides the necessary objects and functions to do Bayesian inference given a probability model and data. It wraps the CmdStan command line interface in a small set of Python classes which provide methods to do analysis and manage the resulting set of model, data, and posterior estimates.

• Installation
  • Conda: install CmdStanPy, CmdStan, C++ toolchain
  • PyPI: install package CmdStanPy
  • GitHub: install from the CmdStanPy repository
  • CmdStan Installation
    • C++ Toolchain Requirements
    • Function install_cmdstan
      • Alternate Linux Architectures
    • DIY Installation
  • Locating the CmdStan installation directory
• User’s Guide
  • Overview
  • “Hello, World!”
    • Fitting a Stan model using the NUTS-HMC sampler
      • The Stan model
      • Data inputs
      • Fitting the model
      • Accessing the results
      • CmdStan utilities: stansummary, diagnose
  • CmdStanPy Workflow
    • Compile the Stan model
    • Assemble input and initialization data
    • Run the CmdStan inference engine
    • Validate, view, export the inference engine outputs
      • Output data
  • Controlling Outputs
    • CSV File Outputs
    • Logging
  • CmdStanPy Examples
    • MCMC Sampling
      • Overview
      • Fitting the model and data
      • Checking the fit
      • Accessing the sampler outputs
      • Saving the sampler output files
      • Parallelization via multi-threaded processing
    • Maximum Likelihood Estimation
    • Variational Inference using Pathfinder
      • Example: variational inference with Pathfinder for model bernoulli.stan
      • Pathfinders as initialization for the MCMC sampler
    • Variational Inference using ADVI
      • Example: variational inference for model bernoulli.stan
    • Using Variational Estimates to Initialize the NUTS-HMC Sampler
      • Model and data
      • Run Stan’s pathfinder or variational algorithm, obtain fitted estimates
    • Generating new quantities of interest.
      • Example: add posterior predictive checks to bernoulli.stan
    • Advanced Topic: Using External C++ Functions
• API Reference
  • Classes
    • CmdStanModel
      • CmdStanModel
    • CmdStanMCMC
      • CmdStanMCMC
    • CmdStanMLE
      • CmdStanMLE
    • CmdStanLaplace
      • CmdStanLaplace
    • CmdStanPathfinder
      • CmdStanPathfinder
    • CmdStanVB
      • CmdStanVB
    • CmdStanGQ
      • CmdStanGQ
  • Functions
    • compile_stan_file
      • compile_stan_file()
    • format_stan_file
      • format_stan_file()
    • show_versions
      • show_versions()
    • cmdstan_path
      • cmdstan_path()
    • install_cmdstan
      • install_cmdstan()
    • rebuild_cmdstan
      • rebuild_cmdstan()
    • set_cmdstan_path
      • set_cmdstan_path()
    • cmdstan_version
      • cmdstan_version()
    • set_make_env
      • set_make_env()
    • from_csv
      • from_csv()
    • write_stan_json
      • write_stan_json()
• What’s New
  • CmdStanPy 1.2.5
  • CmdStanPy 1.2.4
  • CmdStanPy 1.2.3
  • CmdStanPy 1.2.2
  • CmdStanPy 1.2.1
  • CmdStanPy 1.2.0
  • CmdStanPy 1.1.0
  • CmdStanPy 1.0.8
  • CmdStanPy 1.0.7
  • CmdStanPy 1.0.6
  • CmdStanPy 1.0.5
  • CmdStanPy 1.0.4
  • CmdStanPy 1.0.3
  • CmdStanPy 1.0.2
  • CmdStanPy 1.0.1
  • CmdStanPy 1.0.0
• Community
  • Project templates
  • Software

Index

next

Installation