[Rd] Kernlab, ksvm, class.weights, prob.model = TRUE (PR#11971)

renriquez at bolsamadrid.es renriquez at bolsamadrid.es
Thu Aug 7 11:27:29 CEST 2008


Full_Name: Ramón Enríquez Gabeiras
Version: 2.7.0
OS: Windows 32
Submission from: (NULL) (193.110.154.163)


For unbalanced problems, if you provide class.weights parameters (i.e
class.weights = c("1" = 80, "0"= 20), prob.model = TRUE) always return the error
...


If you try with prob.model = FALSE it works perfectly ....

Thanks in advance and regards.



More information about the R-devel mailing list