[R] Fwd: Error in setting the parameter values of k

Patrick (Malone Quantitative) m@|one @end|ng |rom m@|onequ@nt|t@t|ve@com
Tue Dec 29 17:53:41 CET 2020


Likely, yes. Your error message says k must be at least 1, so searching
below 1 is probably your issue.

Also, logically, zero nearest neighbors doesn't seem to make a lot of sense.

Pat

On Tue, Dec 29, 2020 at 11:01 AM Neha gupta <neha.bologna90 using gmail.com>
wrote:

> Thank you for your response.
>
>  Are you certain that k = 0 is a legitimate setting?
>
> Since, the default value of k is 1, I wanted to search between the values
> of 0 to 3.
>
> Milne, Do you mean I have to provide both the lower and upper bounds
> greater than 1 in order to get rid of this error?
>
> On Tue, Dec 29, 2020 at 4:50 PM T. A. Milne <milneta using tuta.io> wrote:
>
> > I am using mlr3 'fast nearest neighbor' leaner i.e. fnnIts parameter is
> > 'k' which has a default value of 1. When I use tuningusing random
> search, I
> > set the parameter of k as: lower= 0, upper=3But it gives an error
> > messageError in self$assert(xs) :  Assertion on 'xs' failed: k: Element 1
> > is not >= 1.I have tried different values but the error remains.Warm
> > regards    [[alternative HTML version deleted]]
> > I know absolutely nothing about the specific statistical tools being used
> > here, but:
> >
> > In several nearest neighbor routines, the parameter k refers to the
> number
> > of nearest neighbors to be considered (in some computation).  In that
> case,
> > k must be at least 1, which is what the cited error message seems to be
> > claiming.  Are you certain that k = 0 is a legitimate setting?
> >
> >
> > - T. Arthur Milne
> >
> >
> >
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help using 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.
>


-- 
Patrick S. Malone, Ph.D., Malone Quantitative
NEW Service Models: http://malonequantitative.com

He/Him/His

	[[alternative HTML version deleted]]



More information about the R-help mailing list