Requirements

1. Python 3.X with pip/Anaconda on Windows
2. Java 1.8 or higher

Install

$ pip install tuo_ml

If you do not have administrator rights append --user/--break-system-packages to above command.

Alternatively, if some errors occur in later steps to install from source.

Run it

tuo/data $ tuo_train --help
tuo/data $ tuo_train

The output should look similar to:

Loading database from database.yml...done
Parsing database...
100%|█████████████████████████████████████| 1705/1705 [00:20<00:00, 84.00it/s]
100%|███████████████████████████████████████████| 1/1 [00:20<00:00, 20.30s/it]
Training...
Win......Average Error: +-1.752% (+-0.027%)
Stall....Average Error: +-0.750% (+-0.017%)
Loss.....Average Error: +-1.225% (+-0.022%)
Points...Average Error: +-1.760 (+-0.031)

The uncertainty of the training is indicated. You can copy the *.pmml files into your data/ folder to use them in tuo now.

Run it from source

If you want to define your own ML model follow these steps.

Download

Download the source zip from github and unzip it. Or better clone the repo

$ git clone https://github.com/APN-Pucky/tyrant_optimize.git tuo
$ cd tuo/ml

Install

tuo/ml $ poetry install 

Modify train.py to your likings.

Run it

tuo/data $ poetry run tuo_train