Multiple Inference documentation

A statistics package for comparing multiple parameters (e.g., multiple treatments, policies, or subgroups).

Statement of need

Researchers often compare many “things” simultaneously. Policymakers regularly compare the effects of many possible policy decisions. Psychologists study individual differences by comparing many people. Diversity researchers compare many demographic groups. Economists compare states and countries in terms of economic variables. Geneticists search for correlations among many genes and phenotypic traits.

We designed this package to help practitioners quickly, easily, and accurately perform such comparisons. It uses a statsmodels-like API and provides template notebooks for ease of use. In just a few clicks, you can upload a .csv file of conventional estimates (e.g., ordinary least squares or instrumental variables estimates) to a Jupyter binder and click “run” to apply a suite of multiple inference tools to your data.

Motivation

Multiple inference techniques outperform standard methods like ordinary least squares (OLS) and instrumental variables (IV) estimation for comparing multiple parameters. This example shows how to apply Bayesian estimators to a randomized control trial testing many interventions to increase vaccination rates.

Start here

The “Multiple Inference Cookbook” allows you to apply a suite of multiple inference tools to your data without writing a single line of code. Click here to see the cookbook, or click the badge below to launch a Jupyter binder and apply the cookbook to your data.

Installation

Install the latest stable build.

$ pip install multiple-inference

Prior to version 1.1.0, the Python package was called conditional-inference.

Install the latest dev build.

$ pip install git+https://gitlab.com/dsbowen/conditional-inference.git

Issues

Please submit issues here.

Contents

• API reference
  • Base classes
  • Distributions
  • Utilities
  • Simultaneous confidence sets
  • Ranking conditions
  • Significance conditions
  • Base classes
  • Improper prior
  • Normal prior
  • Nonparametric prior
• Examples
  • Multiple Inference Cookbook
  • Bayes primer
  • Inference after ranking
  • Distributions
• Contribute
• Changelog
  • 1.2.0
  • 1.1.0
  • 1.0.0
  • 0.0.3
  • 0.0.2
  • 0.0.1

Indices and tables

• Index

• Module Index

• Search Page

Citations

@article{Bowen2022multiple,
   doi = {10.21105/joss.04492},
   url = {https://doi.org/10.21105/joss.04492},
   year = {2022},
   publisher = {The Open Journal},
   volume = {7},
   number = {75},
   pages = {4492},
   author = {Dillon Bowen},
   title = {Multiple Inference: A Python package for comparing multiple parameters},
   journal = {Journal of Open Source Software}
}

Acknowledgements

I would like to thank Isaiah Andrews, Toru Kitagawa, Adam McCloskey, and Jeff Rowley for invaluable feedback on my early drafts.

My issue templates are based on the statsmodels issue templates.