Robust Regression

This is a bunch of scripts to make produce the simulations for the paper, but organized in a handy package that can be expanded later in time.

Code Organization

The package is contained in the folder robust_regression is subdivide into 5 other packages:

• sweeps : contains helpers to create sweeps of some parameters
• aux_functions : contains the definitions of fully vectorized (á la Numpy) function like $\mathcal{Z}{\text{out}}$, $f{\text{out}}$ and $f_{\mathbf{w}}$.
• fixed_point_equations : contains the definitions of the fixed point form for the problems studied. Also it contains the functions to run a single fixed point iteration or to optimize over some parameters.
• regression_numerics : contains the relevant routines to perform ERM simulations and AMP simulations
• utils : miscellaneus stuff used for numerical integration or root finding.

The workspace contains also some resources that can be looked at when setting up a simulation. Inside the directory examples there are scripts to run basic simulations written with the purpose of clarity for someone new to the package. The directory simulations on the other hand contains more advanced scripts to run simulations where the main objective is not clarity or read or use.

The directory tests contains the unit tests for some functions of the package.

Installation

To install the package in the virtual environment of your choice (venv or condaenv) first activate the virtual environment and then in the folder of this package run

$ pip install .

Once installed the package can be imported as:

import robust_regression as rr

Contributing

Fell free to open pull request if you want to add something to the package.