A relative was looking for buying a property but with the inflation, property prices were just crazy and too high for a certain budget. So we decided to have a clear idea of the property prices in general and per neighborhood in Toronto.
We web scraped Zolo and grabbed data for 827 houses, 1757 Condos, and 312 Town-Houses for sales ( each cart has 24 articles). The huge majority of :
- Houses: around 7oo k and 5.5M { Mean: 2.5 M}
- Neighborhood with more properties: Scarborough Village, South Parkdale, Black Creek
- N_ price under the mean: Yonge-St, North Riverdale, Regent Park, Henry Farm, Church Yonge Corridor …
- Most expensive: Niagara
- Condos: around 18o k and 2M { Mean: 1M}
- Neighborhood with more: East York, Elms-Old Rexdale, Mount Dennis
- N_ price under the mean: Most
- Most expensive: Greenwood - Coxwell
- Townhouses: around 7oo k —> 4M {Mean: 1.7M}
- Neighborhood with more: Woburn, Alderwood, Humberlea-Pelmo, Danforth
- N_ price under the mean: Milliken, Hillcrest Village
- Most expensive: Dorset Park
For the web scraping, we used Selenium mostly because Zolo is a dynamic Website.
- We set up a driver with a specific for each kind of property
- Then Sign up through this driver: To sign up, you need to write your email address and then write the access number that they send you at this address. (Sleep function to not being caught)
- Grab the information from each page (24 articles per page) and insert them in a dictionary
- Data Cleaning:
- It wasn’t possible to grab directly the price, number of beds and bathrooms, and the size separately so we define a function for that.
- Made a column call room = Number of bathrooms + number of beds
- Put the size at the same format