Currently, we don't have price variables to include in HLCM and ELCM models. These variables include both residential and non-residential prices. I think @mxndrwgrdnr is taking care of non-residential prices for ELCM model. I'll focus on the residential price for both rentals and sales. One way to generate these vars is estimating Real estate price models for owners and renters and use the predicted price from these models in other models (HLCM and ..). Another alternative is using the observed prices from Craigslist data for rent and Sales data from either DataQuick or Redfin data that we already have in Buildings table.
Since we don't have REPM for bayarea yet, I'll use Craigslist data to apply into the HLCM model for now. Later we can make decisions about using this data or update it for final models.
Tasks for Rentals:
Upload the Craigslists data into UAL box.
Update the datasource.py
Make the required variables
Check the quality of data for created features
Tasks for Sales:
Check the quality of redfin data, use it if it's good. Coordinate with @waddell for sales data from DataQuick
Currently, we don't have price variables to include in HLCM and ELCM models. These variables include both residential and non-residential prices. I think @mxndrwgrdnr is taking care of non-residential prices for ELCM model. I'll focus on the residential price for both rentals and sales. One way to generate these vars is estimating Real estate price models for owners and renters and use the predicted price from these models in other models (HLCM and ..). Another alternative is using the observed prices from Craigslist data for rent and Sales data from either DataQuick or Redfin data that we already have in Buildings table. Since we don't have REPM for bayarea yet, I'll use Craigslist data to apply into the HLCM model for now. Later we can make decisions about using this data or update it for final models.
Tasks for Rentals:
Tasks for Sales: