RaceApp
Flask App for the Big Data Coding Competition (it-talents)
- Data Visualisation with DC,D3 and Crossfilter
- Python 2.7 is used
- Used packages can be found in requirements.txt
- Predictive Model is built and tuned in advance with sklearn. Random Forest is compared with logistic Regression. Logistic Regression is chosen in final. Predictive Model is dumped in a file and loaded each time when it is used.
- Analogous to a Data Warehouse with OLAP functionalities, sql queries are saved as new tables to reduce calculation time --> Break of ACID philosopy
Principal information
- There are two possible Login Roles.
- Admin-Role:
- Login: admin@admin.com
- Password: admin2016
- User-Role:
- Login: 1@user.com
- Password: 123456
- Admin-Role:
Installation
-
Be sure to have mysql installed
-
Use following codes in Bash/Terminal
git clone
mysql < schema.sql
pip install -r requirements.txt
export FLASK_CONFIG=development
export FLASK_APP=run.py
flask run
Further improvements
- Use Docker (Docker-Compose) for easier installation and use
- Create single executable
- Adding races
- Therefor building stored procedures in mysql for the calculated tables