Setup the Code

git clone git@github.com:MarcusBluestone/Foulball_DecisionTree.git
conda env create --name foulball -f environment.yml
conda activate foulball

Data Setup

1. Ignore the playground and scraper folders; they are irrelevant
2. Go into src/mainand change the data_path variable to the filepath location of your data.
3. Updated the feature_types and output_col variables to match precisely the relevant columns in your data (note: feature_types is a dictionary mapping column_name -> 'num'/'one-hot'/'ord')

Run the Code

You can change the parameters in src/main (max_depth, top_k, random_state, test_size) however you'd like.

Run the code python -m src.main

It will output test accuracy, and a legible version of the Tree Classifier