Metadata-Version: 2.1
Name: kwnpeb
Version: 0.1.11
Summary: Compute the Kiefer-Wolfowitz nonparametric maximum likelihood estimator
Home-page: https://github.com/sit836/KW_NPEB
Author: Sile Tao, Li Zhang, Guanqi Huang
Author-email: sile@ualberta.ca, lzhang2@ualberta.ca, frank.huangguanqi@gmail.com
License: UNKNOWN
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Description-Content-Type: text/markdown
Requires-Dist: scipy (==0.19.1)

# Kiefer-Wolfowitz Nonparametric Empirical Bayes
Compute the Kiefer-Wolfowitz nonparametric maximum likelihood estimator for mixtures. 

In contrast to the previous approaches, the optimization problem is reformulated into a convex problem by 
[Koenker and Mizera (2014)](http://www.stat.ualberta.ca/~mizera/Preprints/brown.pdf)'s method and efficiently solved by 
interior-point method.

## Making Predictions With No Features - A Basic Usage
Given a training set T = {y_i}, the algorithm provides a way to construct a predictor of future y-values such that the sum 
of squared errors between observations and predictors is minimized.  

## Getting Started
### Prerequisites 
You will need:
* python (>= 3.6)
* pip (>= 19.0.3)
* MOSEK (>=8.1.30) 

Important about MOSEK:
* MOSEK is a commercial optimization software. Please visit [MOSEK](https://www.mosek.com/) for license information.
* PIP: 
```
pip install -f https://download.mosek.com/stable/wheel/index.html Mosek --user
``` 
For different ways of installation, please visit their [installation page](https://docs.mosek.com/8.1/pythonapi/install-interface.html).
* MOSEK needs to be installed in the GLOBAL environment. 



### Installing
```
pip install kwnpeb
```

## Examples
* [simple](https://github.com/sit836/KW_NPEB/tree/master/examples/simple) - The basic usage
* [bayesball](https://github.com/sit836/KW_NPEB/tree/master/examples/bayesball) - In-season prediction of batting averages with the 2005 Major
League baseball

## Contributors
* [Sile Tao](https://ca.linkedin.com/in/sile-tao-95523941)
* [Li Zhang](https://ca.linkedin.com/in/li-zhang-0350833b)
* [Guanqi Huang](https://ca.linkedin.com/in/guanqi-huang)

## License
This project is licensed under the MIT License - see the [LICENSE.md](LICENSE.md) file for details


