If your data set contains primarily individuals residing in Canada, you can create a derived indicator variable in Data Desk to identify who does not live in an urban area or small town. Just one more potential predictor variable to test in your models.
First, your data set needs to contain Postal Code. You’ll also need to include Country, if your data set includes non-Canadian addresses.
The expression below will create an indicator variable with a value of ‘1’ (true) for rural Canadian addresses, and ‘0’ (false) for urban addresses.
Drag your own variables for Postal Code and Country into the expression window.
mid('Postal Code',2,1) = 0 AND textof('Country') = "Canada"
CoolData blog 