FSjoanlouie

Regression Analysis - Vector and Raster data

Discussion created by FSjoanlouie on May 8, 2014
I am prepping the spatial datasets to run a regression model.  The statistics have already been calculated outside of ArcGIS and I have the coefficients for each of my variables.  I will be running the model via Raster Calculator/map algebra and plugging in the regression equation.

e.g.
output = 1.398 + (-0.432 * ELEV) + (0.231 * LAT) + (0.381 * SOIL_A) + (-0.1112 * SOIL_B)


What I am having trouble with is the vector (polygon) dataset for soil type.  I have the coefficient values for each of the soil types.  Do I create another field in the polygon dataset for the coefficients, populate each of the soil types with their respective coefficients, then extract out by soil type to convert to separate raster layers?  The resulting raster layer cells would all have the same [coefficient] values (per soil type).  With those separate soil raster layers, are they added to the model as is?

so instead of the above equation, would it actually be:

output = 1.398 + (-0.432 * ELEV) + (0.231 * LAT) + SOIL_A + SOIL_B

for my particular datasets? (I know it's probably not the correct syntax for raster calculator, but my question is about the general process).

Or is this completely wrong and I need to be doing something else?

Thanks, this seems like it should be very easy but I can't find any reference to this (or I'm not searching the appropriate terms/concepts)

Outcomes