I am noticing that there is a difference between the fitted.values returned by train.kknn, and the values returned using predict with the same model and dataset. For example:
> data (glass) > tmp <- train.kknn(Type ~ ., glass, kmax=1, kernel="rectangular", > distance=1) > tmp$fitted.values [[1]] [1] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 [62] 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 1 2 2 1 2 2 5 2 2 2 6 2 2 2 2 2 2 2 2 2 2 2 2 [123] 2 2 2 2 3 2 2 2 5 5 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 2 3 3 3 3 3 3 2 3 7 5 5 5 5 5 5 5 5 5 5 2 5 6 6 6 6 6 6 6 [184] 2 6 7 7 2 6 7 7 7 7 7 7 7 7 7 7 7 7 5 7 7 7 7 7 7 7 7 7 7 7 7 attr(,"kernel") [1] rectangular attr(,"k") [1] 1 Levels: 1 2 3 5 6 7 > predict (tmp,glass) [1] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 [62] 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 [123] 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 5 5 5 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 [184] 6 6 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Levels: 1 2 3 5 6 7 When I check the confusion matricies for these I see that fitted.values is giving some confusion, that is, like it is a true fit, whereas predict is returning the exact answers. > table (tmp$fitted.values[[1]],glass$Type) 1 2 3 5 6 7 1 69 4 0 0 0 0 2 1 67 2 1 1 1 3 0 1 15 0 0 0 5 0 3 0 11 0 1 6 0 1 0 0 8 1 7 0 0 0 1 0 26 > table (predict(tmp,glass),glass$Type) 1 2 3 5 6 7 1 70 0 0 0 0 0 2 0 76 0 0 0 0 3 0 0 17 0 0 0 5 0 0 0 13 0 0 6 0 0 0 0 9 0 7 0 0 0 0 0 29 Can anyone clarify what fitted.values and predict actually do? I would have expected they would give the same output. Thanks... Jonathan -- View this message in context: http://r.789695.n4.nabble.com/kknn-predict-and-kknn-fitted-values-tp4711625.html Sent from the R help mailing list archive at Nabble.com. ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.