Day 58

Aug 26, 2017 · 197 words

As my final work week approaches, it’s becoming more apparent that we might not have enough time to fully integrate the prediction system with the web app unless other people are able to keep working on it during the next few weeks.

As a precaution I want to ensure that I have enough documentation of the system I have in mind. Today I wrote a README guide for my GitHub repo that generally goes over how to use the modeling system and a bit about how it works. There’s also plenty of documentation within the actual module.

I also made a demo folder that contains a sample prediction environment. It holds a prediction script that showcases energize.MultiRFModel in action along with some simple Pandas usage. There’s some clean sample input and output files using the naming convention I drafted yesterday.

Lastly, I mentioned yesterday that you should be able to select certain columns from the input data to model and predict. That was fairly simple to implement. It can be used by passing a list of column names to the new columns parameter of MultiRFModel. The default option of None will use all data columns.

comments powered by Disqus