Skip to content
Prev 6609 / 29559 Next

Simulating Two Surfaces with a user-defined Cross-Correlation Index

If the variables are spatially uncorrelated, it is a special case of
what I provided, and you could specify the variography in terms of pure
nuggets, and/or use a neighbourhood of 1 (or 0?).

As an alternative, fill the grids with data generated with function
rmvnorm in package mvtnorm.

I find it however hard to understand what you mean by a spatial
cross-correlation index, especially the word spatial, when you do not
want to address spatial autocorrelation of the variables themselves.
--
Edzer
Enrico R. Crema wrote: