This project contains a solution for Kaggle Grupo Bimbo challenge. The solution itself is not designed to get a good score (for that I have a complete solution, not really pretty one), but to try out if Luigi can be a good tool for competitions.

For running tasks I use Luigi. It's great for making tasks granular so that you dump all the results after the task which guarantees you won't struggle with memory leaks.

I'm planning to work on this more, both on Luigi plumbing and feature generation. Not sure when though.

You can use any python distribution, whether its dockerized or not.

Here's what I'm doing with Anaconda installation:

~/anaconda/bin/python -m venv . --without-pip --system-site-packages curl https://bootstrap.pypa.io/get-pip.py | ./bin/python source ./bin/activate