franchesoni / nasa_space_apps_lrmc

Low Rank Matrix Completion by Proximal Gradient

Feeling the data!

Implementation of Low Rank Matrix Completion by Proximal Gradient, and data visualization using t-distributed Stochastic Neighbor Embedding. There is an example of matrix completion where YaleB-Dataset (of faces) is used. Visualization examples use "Near Earth Comets" and "Meteorite_Landings" datasets, which were provided by NASA as example resources on SpaceApps challenge "Chasers of the lost Data".

Dependencies

• numpy
• pandas
• sklearn
• plotly
• matplotlib

Demo

The implemented function for Low Rank Matrix Completion is in

• lrmc.py and uses auxiliary functions that are in
• aux_functions.py

The script that runs the function and establishes the parameters is:

• lrmc_test.py

The code will use YaleB-Dataset.zip uncompressed in this directory.

The script that generates the output images is

• vizualization.py

This can also be ran from the interactive notebook called "FEELING DATA.ipynb"