On 9 May 2013 02:35, jeremy theler <[email protected]> wrote: > what I want to do is to > interpolate arbritarily-scattered data, not only points located in a > rectangular grid.
Depending on how statistical you want to be, gaussian processes/fields can be very useful for this sort of thing. GSL helps a bit, but is missing the multivariate normal distribution, but there is code here to extend it: http://www.mail-archive.com/[email protected]/msg00631.html The computationally intensive part is generating the covariance matrix, which is O(n^2) so only really usable when you have less than a few thousand points. Hope that helps! Sam
