README.md

Alvin Smart Money Classification

My solution for the Alvin Smart Money Classification challenge

Requirements

• Python 3.8 or greater
• Pipenv or conda package manager

Usage

1. Create virtual environment and install dependencies

   # pipenv
   pipenv install --dev
   
   # conda
   conda env create -f environment.yml

2. Activate the virtual environment

   # pipenv
   pipenv shell
   
   # conda
   conda activate alvin-smcc

3. Download the data and extract it in the input directory

   # extracting the data in linux
   unzip "input/*.zip"

4. Train a model

   # view train options
   python src/cli.py train --help
   
   # train a model
   python src/cli.py train --model [model] --preprocessor [preprocessor]

5. Make predictions on test data using the trained model. Predictions are saved in the output/predictions directory

   # view predict options
   python src/cli.py predict --help
   
   # obtain predictions
   python src/cli.py predict --run-id [run_id]

Notes

• By default, there are 2 folds for cross validation, but that can be changed with the NUM_FOLDS environment variable
• Verbosity can be changed with the VERBOSITY environment variable
• Environment variables can be set in the .env file