A Python package to conduct regularized linear regression via Lasso, Ridge Regression and Elastic Net, implemented basically by numpy and acceleration including multi-processing, Cython and C++ code wrapped by Cython.
This is a final project product from the course STA-663-2016 by Cliburn Chan [https://github.com/cliburn] and Janice McCarthy at Duke University.
The final project: https://github.com/lguirola/sta663-Final-Project
We survived the course and learnt a lot!
Luis Guirola, lguirola @ Github
Xiaodong Zhai, shldngzh @ Github
This package is not published as a Python site-package yet, so the installation is quite straightforward since you just need to download the package and put the folder where you need, and to make sure you can simply import the package is pretty enought.
Whatever OS you use, just download and import and use.
- C++ compiler: required. Since we use g++ compiler, I believe it would be stable in different platforms.
- Python: 3.5 is what we used in development.
- Pyton site-packages: numpy
- other library/packages: not required.