Install Dependencies

pip install -r requirements.txt
playwright install

Web Scraping

To scrape data using the url provided in main.py, the url is provided for everything except for the pages you desire. Edit the loop for the number of pages you would like to scrape.

Training the model

To train the model, run the cells in the model_reg.ipynb (regression; exact price predictions) or model_cat.ipynb (classification into price buckets) notebook. The notebook will save the model(s) in the models directory.

Running the Flask app

To run the Flask app, run the following command in the terminal (in your virtual environment):

flask run

The app will be available at http://127.0.0.1:5000 (or whatever URL is returned in the URL), where an image can be uploaded to get a price prediction.

Note: the model is saved in Git Large File Storage. You must have it installed to load the pre-trained model.

Generating, publishing the paper

In the main branch, run the following commands to generate the paper:

quarto publish gh-pages paper.qmd