I couldn't find on ?gstat, ?predict.gstat nor here the answers to these little doubts:
:: Having to build a gstat object with a SpatialPixelsDataFrame and a SpatialPointsDataFrame, does the gstat object automatically recognize the supoprt size of the two variables? (That is a block of same dimensions of the cell size in the SPixelsDF, and a point for the SPointsDF instead)
:: Having then to predict a block co-kriged map on an output grid (again a SpatialPixelsDataFrame object) with its own cell size, do I have to explicitly specify the block or does the predict.gstat method automatically compute block co-kriging?
I think I know the answers but I would feel better with 100% certainty.
Thank you all,