Metadata-Version: 2.1
Name: shifterator
Version: 0.3.0
Summary: Interpretable data visualizations for understanding how texts differ at the word level
Home-page: https://github.com/ryanjgallagher/shifterator
Author: Ryan J. Gallagher
Author-email: gallagher.r@northeastern.edu
License: UNKNOWN
Keywords: natural language processing,sentiment analysis,information theory,computational social socience,digital humanities,text analysis,text as data,data visualization,data viz
Platform: UNKNOWN
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Operating System :: OS Independent
Requires-Python: ~=3.6
Description-Content-Type: text/markdown
License-File: LICENSE.txt

# Shifterator

The Shifterator package provides functionality for constructing **word shift graphs**, vertical bart charts that quantify *which* words contribute to a pairwise difference between two texts and *how* they contribute. By allowing you to look at changes in how words are used, word shifts help you to conduct analyses of sentiment, entropy, and divergence that are fundamentally more interpretable.

<p align="center">
  <img src ="docs/figs/shift_sentiment_detailed_full.png" width="400"/>
</p>


## Install

Python code to produce shift graphs can be downloaded via pip.

`pip install shifterator`

## Documentation

[The documentation](https://shifterator.readthedocs.io/en/latest/) details how to create various kinds of word shift graphs with Shifterator, and includes a detailed cookbook for how to interpret, visualize, and work with word shifts.

See the following paper for more details on word shifts, and please cite it if you use them in your work:

> Gallagher, R. J., Frank, M. R., Mitchell, Lewis, Schwartz, A. J., Reagan, A. J., Danforth, C. M., Dodds, P. S. (2021). [Generalized Word Shift Graphs: A Method for Visualizing and Explaining Pairwise Comparisons Between Texts](https://epjdatascience.springeropen.com/articles/10.1140/epjds/s13688-021-00260-3). *EPJ Data Science*, 10(4).


