Hi all,
I'm trying to compute adaptive kernel density estimation in R (multivariate
data), using the k-nearest neighbour method for selecting the bandwidth.
I found locfit (only for univariate) and I tried the Kernsur in GenKern
package (which says it can use non equally spaced and adaptive bandwidths)
but I don't know how the bandwidth slection works in this case.
Any suggestion is welcome.
Thanks
Sandra Oliveira
[[alternative HTML version deleted]]
_______________________________________________
R-sig-Geo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
