pyjags-arviz 0.0.3

Makes MCMC samples from PyJAGS usable in ArviZ

pyjags_arviz

Makes MCMC samples from PyJAGS usable in ArviZ

Table of Contents

1. Installation
2. Getting Started

Installation

1. Install via PIP:

   pip install pyjags_arviz 
   

   or

   pip3 install pyjags_arviz 
   

   if not using Anaconda.

   To get the latest version, clone the repository from github, open a terminal/command prompt, navigate to the root folder and install via

   pip install .
   

   or

   pip3 install . 
   

   if not using Anaconda.

Usage

Import the function convert_pyjags_samples_dict_to_arviz_inference_data via

from pyjags_arviz import convert_pyjags_samples_dict_to_arviz_inference_data

Having sampled the from the posterior distribution using PyJAGS via

samples \
    = jags_model.sample(...)

one can write

idata = convert_pyjags_samples_dict_to_arviz_inference_data(samples)

to convert the dictionary returned from PyJAGS to an ArviZ InferenceData object.

This object can be used in ArviZ to generate trace plots and compute diagnostics.
Trace plot:

az.plot_trace(idata);

Effective sample size:

az.ess(idata)

Gelman and Rubin statistic:

az.rhat(idata)

Autocorrelation plot:

az.plot_autocorr(idata);