[R-sig-Geo] Interpolating from SpatialPointsDataFrame - estimates for new boundaries

Juta Kawalerowicz juta.kawalerowicz at stx.ox.ac.uk
Mon Jan 30 03:06:54 CET 2012


Dear List,

I have the following problem, maybe someone can suggest how to proceed.

I have a SpatialPointsDataFrame for United Kindgdom districts in 2001 
and 1971, both containing number of people in each polygon. I would like 
to calculate population growth between 1971 and 2001. My idea was to use 
Chris Briunsdon's pycnophylactic interpolation package pycno (softens 
the sharp changes at block boundaries) for 1971 and then to get 
estimates of 1971 population but within 2001 boundaries (there is a 
function estimate.pycno which for this in pycno package). It worked 
nicely, but my estimates are a bit strange (i.e do not support a 
hypothesis which I strongly expected to be supported...) and I just 
wondered if I could cross-validate using some other method. How would 
you approach? Is Kriging a way to go?

Thanks,
Juta



More information about the R-sig-Geo mailing list