Use GWR to predict on out of sample data

Discussion created by cml1227 on Dec 6, 2016

I have a set of polygons (~700) that comprise a country and want to use GWR to forecast incidences of violence against civilians. Is it possible to use a subset of the polygons to build the model and then somehow feed in the second subset (that hasn't been used to build the model) and make predictions on those model? Basically, I want to see how good my GWR model is on out-of-sample predictions.