Hi there,
I wonder how can I use OLS (Ordinary least squares) & GWR (geographic weighted regression). I am interested to analyse the fire density (dependent variable) with elevation (independent variable) in my study area. My hypothesis is elevation is a determining factor in forest fires. My data is 250 m resolution.
Finding no way to use raster data in OLS & GWR, I converted it to point shapefiles (millions of data points). I took a sample of it about 600000 data points & run OLS & GWR analysis. It worked but takes a lot of computational time. I am not sure if point shapefiles are appropriate format particularly for GWR.
Is there any suggestion that given the size and format (raster) of my data how can I efficiently run OLS & GWR.
Many thanks.
Muhammad