Mistake in predict.train.kknn

[rkangr]

In some settings it is useful to control scaling variables from outside of kknn. Unfortunately, when train.kknn is used with scale=FALSE, the corresponding predict method does not pass this parameter to kknn used inside predict.train.kknn, so predictions are still computed with default scaling. It seems, that a possible solution would be to add scale component to the value of train.kknn and use that component in predict.train.kknn

Could this mistake be corrected?