Metadata-Version: 2.1
Name: xifa
Version: 0.1.0
Summary: Accelerated Item Factor Analysis
Home-page: https://github.com/psyphh/xifa
Author: Po-Hsien Huang
Author-email: psyphh@gmail.com
License: UNKNOWN
Project-URL: Bug Tracker, https://github.com/psyphh/xifa/issues
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.6
Description-Content-Type: text/markdown
License-File: LICENSE

# **XIFA**: Accelerated Item Factor Analysis
`xifa` is a python package for conducting item factor analysis (IFA). 

`xifa` is build on [`jax`](https://github.com/google/jax), a package for Autograd and XLA. Hence, `xifa` can run IFA on GPUs and TPUs to speed up the training process. That is why we call it *Accelerated IFA*.

In the current version (0.0.1), `xifa` supports ordinal data IFA with graded response model (GRM; [Semejima, 1969](https://link.springer.com/article/10.1007%2FBF03372160)) and generalized partial credit model (GPCM; [Muraki, 1992](https://doi.org/10.1177/014662169201600206)). 

For a tutorial, please see [Big Five 50 Items Example](http://colab.research.google.com/github/psyphh/xifa/blob/master/examples/big5.ipynb).




