I currently have a shape file with approximately 20 numeric variables. Several of these variables have missing values.
As this is a shapefile I do not think using the median or mean as a form of imputation would be effective.
Is there an approach I can take which would look at the specific polygon with a missing value, and impute that polygon with the mean/median of the surrounding polygons?
I have found the following paper:
Baker et al.: Missing in space: an evaluation of imputation methods for missing data in spatial analysis of risk factors for type II diabetes. International Journal of Health Geographics 2014 13:47
which mentions a CAR prior distribution technique- but I am unsure how to implement this technique.
Is there a simplified approach to this spatial imputation technique in r?