kknn
kknn copied to clipboard
KlausVigo
Weighted k-Nearest Neighbors
Fixes #16 Rather than using `options` to set the contrasts, this sets them directly on `model.matrix` via the argument `contrasts.arg`. Feedback/ comments are welcome!
Hi, thanks for developing this package first and foremost! I'd like to have confidence intervals based on the weighted means of the k nearest neighbors for my continuous response variables...
Are you planning to update the `kknn` version on CRAN, or is the GitHub version in some way the official release? Currently the invocation ```r replicate(100, kknn::kknn(speed ~ dist, cars[1:3,...
predict.train.kknn() does not respect all parameters passed to train.kknn(). An example is `scale`. For example, predicting with `scale = FALSE` and `scale = TRUE` with `train.kknn()` give the same results:...
Hello! I think there is a bug where if the testset does not contain levels of a character variable that are in the trainset, the model will not run. ```...
Hello, First of all , thank you very much for your work on this package. I just tried to use it and wanted to apply action to NA values but...
`train.kknn` protects against `kmax` being too large but not `ks`: ``` r library(kknn) train.kknn(mpg ~ ., data = mtcars, ks = 32) #> Error in D[, 1:j]: subscript out of...
When `kknn` is referenced by namespace (as opposed to being fully attached via `library`) an error occurs: ``` r train.con Error in get(ctr, mode = "function", envir = parent.frame()): object...
Here is a data set on which `train.knn()` fails: ``` args dmtmp$cl [1] 0 1 2 2 2 0 0 0 [9] 1 2 1 1 32676 32676 32676 32676...
