binomialbias 1.3.3
Last updated:
0 purchases
Description:
binomialbias 1.3.3
BinomialBias
This library computes and plots quantitative assessments of discrimination within organizations, based on the binomial distribution.
This code supports the following paper:
Quantitative measures of discrimination with application to appointment processes. Robinson PA, Kerr CC (2024). PLoS ONE 19(3): e0299870. https://doi.org/10.1371/journal.pone.0299870
There are several ways to use this library, described below.
Webapp
A live webapp is running at https://binomialbias.sciris.org.
Local installation and usage
Python
To use locally with Python, run
pip install binomialbias
This can then be run via e.g.:
import binomialbias as bb
bb.plot_bias(n=20, n_e=10, n_a=7)
This example shows the statistics for the case where there were n = 20 appointments (e.g., the size of a committee), out of which n_e = 10 people were expected to belong to a given group (e.g., female), and for which n_a = 7 actually were.
Shiny
To run the Shiny app locally, clone the repository from GitHub, then install with
pip install -e .[app]
The Shiny app can then be run locally via the run script.
Structure
All code for the Python package is in the binomialbias folder.
The script for generating the figure in the paper is in the scripts folder.
Continuous integration tests are in the tests folder.
Older Jupyter and Matplotlib versions are available in the archive folder.
License:
For personal and professional use. You cannot resell or redistribute these repositories in their original state.
Customer Reviews
There are no reviews.
