Cataclop

horse race supervised learning and autonomous betting

EXPLORATION

• exploration
• onboarding

1/ INSTALL

cp .env.dist .env

docker-compose up -d

2/ OBTAIN THE JSON RACES FILES

• option 1 : extract an existing archive

tar xzvf var/archives/races-2021-01.tgz -C var/scrap

• option 2 : scrapping

eg. scrap all January 2021 races

docker exec -it cataclop_app python manage.py scrap "2021-01-01" "2021-01-31"

3/ IMPORT THE JSON FILES

eg. import Janurary 2021 races

docker exec -it cataclop_app pipenv run python manage.py parse "2021-01-*"

4/ TRAIN MODELS

docker exec -it cataclop_app python manage.py shell_plus --notebook

open the output link (the one starting with http://127.0.0.1:8888/) in your navigator to access the jupyter notebooks

open the exploration notebook to get an overview of the data open the onboarding notebook for a quick start on creating models

REST API

docker exec -it cataclop_app python manage.py createsuperuser

The REST API should be running on http://127.0.0.1:8082/api and admin dashboard on http://127.0.0.1:8082/admin

(advanced) BETTING

docker exec -it cataclop_app python manage.py bet --loop

Tests

docker exec -it cataclop_app python manage.py test --exclude-tag=integration