ns knn function

[PierreScl]

Hey,

As you are transposing data_miss before computing the distances I understand that samples should be in columns of data_miss (It is consitent with the computation of minimums of the rows in minVals = apply(data_miss, 1, min, na.rm=TRUE) ).
However, in the scaling part at the very beginning the scale function is applied on non-transposed data_miss thus on the samples.

I was wondering if it was not an error ? Because when we use the scale argument of the function, the results we get are weird.

Thanks,

Pierre