Toronto House Price Study

A preliminary data driven study on Toronto house prices

1) Install requirements.txt for python packages used in the study

2) Toronto Real Estate Board Home Price data is extracted from pdf from Feb 2012 to Oct 2020

To scrap the pdfs from TREB website:

• run utils/extract_housing_price.py

To extract home price data from pdfs:

• first run utils/convert_housing_pdf_csv.py (uses tabula-py package)
• then run utils/read_housing_csv.py -> creates csv file located in /data/treb_home_price_reference_master/TREB_Home_Price_Reference_2012Feb_2016Oct.csv
• Analysis notebook toronto-house-prices.ipynb is run referencing csv files in data folder
• web scraping and pdf->csv generation do not need to be run but source files are provided in /data for reference and utils scripts are provided in /utils
• sources of data in /data are provided in toronto_house_price_study.pdf

3) Model uses Prophet package (fbprophet)

• requires first pystan installed and configured
• if issues are encountered installing Prophet see https://facebook.github.io/prophet/docs/installation.html#python