• This repository contains the code developed during my internship at LightOn
• Version 0.1

To run the code, the following python libraries are needed:

• numpy
• scipy
• sklearn
• matplotlib
• time
• os

To recover the plots of my internship final report, run the correspondent scripts. Example: python script/QR_scripts.py

Some scripts require real datasets, download them at:

• SalinasA and Indian Pines datasets for NMF
• ORL and Yale faces datasets for k-means

and put them in a directory called dataset.

In the core directory, two tipes of files are present:

• main files contain basic functions to run the new algorithms (for example: QR.py). If you just want to use the algorithm, only these are the files that you need
• tools files that perform analysis of the correspondent algorithms and plot figures (for example QR_tools.py)

This core files are imported by the correspondent scripts files when executed.