Webscraping of blocket.se in the search for a new car...
First of all, don't fly when you can drive and don't drive when you can walk or bike, as the global temperature is rising! But if you like me really need a car, in my case to do longer trips instead of flying, here is a story about how I selected the used car I decided to buy.
You need Python3, and GIT.
- Clone this repo.
- Install the requirements: pip install -r requirements.txt
Write a script similar to run_updata.py, where you specify the blocket search URL:s that you are interested in. The Result will be saved in a csv file.
Run a regression similar to the Jupyter Notebook: regression.ipynb. This fits a model to the data to predict the price of the cars. The cars are then rated based on how much cheaper or more expensive they are compared to this model.