Project Design and Overview
The goal of this project is to predict the price of the Toyota 86/Subaru BRZ as well as tracking the market conditions using statistics, analytics. This project will be used to help me purchase my attainable dream car in the near future. Listed below are the main components of the project.
-
1) Python ETL scheduled using Cron
- Data is collected using a python web scrapping script hosted on GCP set to run hourly.
- Data is validated/cleaned and inserted into Heroku Postgres database.
-
2) Car Data REST API
-
3) Regression Model
- Data is cleaned and tested for Covariance.
- A reduced model is built using forwards selection.
- Daily script runs to create coefficients for the chosen reduced model and is stored in database.
- Code for Regression Model
-
4) Flask App
- Website is built on Flask web framework and bootstrap.
- Car Price Estimater inputs user parameter and price is created using latest daily generated regression model.
-
5) Google Data Studio Dashboard
- Market Trends are displayed using live data from Heroku Postgres database.
- Data collection statistics are displayed at bottom
-
6) New Listing Alerts
- Python script scans database daily to look for new listings.
- New listings and important listing information are sent via email alerts along with listing url
-
5) Github
- Github Repo contains all code for Flask app, Scripts, DB sql create, and statistical model creation.
*Car Price Estimator, Dashboard, and Regression Model is located under the drop down in the navigation bar.
<