GitHub - quantregpy/quantregpy: A translation of the popular R library quantreg to python for quantile regression

Update September 2021

Scikit Learn introduced version 1.0, and with it an implementation of quantile regression. Granted, there are some features that exist here that don't exist in Scikit Learn, it doesn't make much sense to keep working on this project in parallel. Development on this project is therefore suspended.

A translation of the popular R library quantreg (Authored by Roger Koenker) to python for quantile regression

**This library is very much experimental at this point.

Fortran code from the quantreg R package is modified to work with f2py

R code from the quantreg R package is translated into python, making heavy use of numpy.

The quantreg R package was accessed through github and distributed by GPL. I cannot overstate my gratitude for the original authors, some of whom are referenced as follows:

Koenker, R. (2017). Quantile regression in R: a vignette. Version of April, 17(2017), 88.

Esmond G. Ng and Barry W. Peyton, "Block sparse Cholesky algorithms on advanced uniprocessor computers". SIAM J. Sci. Stat. Comput. 14 (1993), pp. 1034-1056.

John R. Gilbert, Esmond G. Ng, and Barry W. Peyton, "An efficient algorithm to compute row and column counts for sparse Cholesky factorization". SIAM J. Matrix Anal. Appl. 15 (1994), pp. 1075-1091.

Koenker, R. (2005). Quantile Regression (Econometric Society Monographs). Cambridge: Cambridge University Press. doi:10.1017/CBO9780511754098

Koenker, R., Chernozhukov, V., He, X., & Peng, L. (Eds.). (2017). Handbook of quantile regression.

If this project matures, I will write better documentation. However, I am trying to mimic the R package, so their documentation should be useful

Installation

Dependencies: For windows and Linux, pip install works fine. For Macs, you may have some trouble, and will need a FORTRAN compiler.

Sample usage

>>> import numpy as np
>>> from quantregpy.quantreg import rq
>>> import pandas as pd
>>> df = pd.DataFrame(dict(a=[1,2,3,4], y=[2,3,4,5]))
>>> out = rq("y ~ a", np.array([.5]), df)
>>> out
{'coefficients': array([1., 1.]), ... etc