To help you get started, we've collected the most common ways that pystan is being used within popular public projects.

IBM / yaps / tests / stan / test_stan2yaps.py View on Github

def run_test(dir):
    pathlist = Path(dir).glob('*.stan')
    nb_test = 0
    nb_success = 0
    nb_bad = 0
    for p in pathlist:
        path = str(p)
        with open(path, 'r') as fin:
            code = fin.read()
            try:
                pystan.stanc(model_code=code)
                nb_test += 1
                try:
                    source = yaps.from_stan(code_file=path)
                    ast_ = yaps.from_string(source)
                    stan = yaps.to_stan(ast_)
                    pystan.stanc(model_code=stan)
                    nb_success += 1
                except (AttributeError, SyntaxError, TypeError, AssertionError, ValueError) as err:
                    print("FAILED\t", path, err)
                except RuntimeError:
                    nb_success += 1
            except (ValueError, RuntimeError):
                nb_bad += 1

    print("-------------------------")
    print("{}% of success ({}/{} valid Stan examples, {} bad examples ignored)".format(

View more ways to use pystan

PyStan

PyStan is a Python interface to Stan, a package for Bayesian inference.

Stan® is a state-of-the-art platform for statistical modeling and high-performance statistical computation. Thousands of users rely on Stan for statistical modeling, data analysis, and prediction in the social, biological, and physical sciences, engineering, and business.

Notable features of PyStan include:

• Automatic caching of compiled Stan models
• Automatic caching of samples from Stan models
• An interface similar to that of RStan
• Open source software: ISC License

Getting started

Install PyStan with pip install pystan. PyStan runs on Linux and macOS. You will also need a C++ compiler such as gcc ≥9.0 or clang ≥10.0.

The following block of code shows how to use PyStan with a model which studied coaching effects across eight schools (see Section 5.5 of Gelman et al (2003)). This hierarchical model is often called the "eight schools" model.

import stan

schools_code = """
data {
  int J;         // number of schools
  array[J] real y;              // estimated treatment effects
  array[J] real sigma; // standard error of effect estimates
}
parameters {
  real mu;                // population treatment effect
  real tau;      // standard deviation in treatment effects
  vector[J] eta;          // unscaled deviation from mu by school
}
transformed parameters {
  vector[J] theta = mu + tau * eta;        // school treatment effects
}
model {
  target += normal_lpdf(eta | 0, 1);       // prior log-density
  target += normal_lpdf(y | theta, sigma); // log-likelihood
}
"""

schools_data = {"J": 8,
                "y": [28,  8, -3,  7, -1,  1, 18, 12],
                "sigma": [15, 10, 16, 11,  9, 11, 10, 18]}

posterior = stan.build(schools_code, data=schools_data)
fit = posterior.sample(num_chains=4, num_samples=1000)
eta = fit["eta"]  # array with shape (8, 4000)
df = fit.to_frame()  # pandas `DataFrame`

Citation

We appreciate citations as they let us discover what people have been doing with the software. Citations also provide evidence of use which can help in obtaining grant funding.

To cite PyStan in publications use:

Riddell, A., Hartikainen, A., & Carter, M. (2021). PyStan (3.0.0).

Or use the following BibTeX entry:

@misc{pystan,
  title = {pystan (3.0.0)},
  author = {Riddell, Allen and Hartikainen, Ari and Carter, Matthew},
  year = {2021},
  month = mar,
  howpublished = {PyPI}
}

Please also cite Stan.