Skip to content
Prev 263226 / 398502 Next

(no subject)

Hi Ungku, your error is in your call to lda.default; the object that
you want to plot, lda_result, is thus not created and cannot be
plotted.

The error you're getting with LDA doesn't look to be a coding error,
but a statistical one. I don't know anything about LDA, but simply
googling "variablest constant within results" came up with several
useful links that explain the problem. Have you looked at these? The
first I found was this:
http://miner.chem.purdue.edu/Lectures/Lecture10.pdf , which explains
the related theory behind this error, and
http://r.789695.n4.nabble.com/Linear-Discriminant-Analysis-error-quot-Variables-appear-constant-quot-td3527288.html

The second link, by the way, is an example of an R question with a
clear subject line and description of the problem, making it much
easier for us to help.

I notice that you have already asked questions related to this same
function and error message in this list, and have received some
suggestions from people who know about LDA. Have you tried these
suggestions? I don't see any trace of them in your code.

Best,
Ethan

P.S. Please include the r-help list in the to: field in any follow-up
emails so other people can chime in or learn from this exchange.
On Mon, Jun 20, 2011 at 7:50 PM, Ungku Akashah <kaslah90 at yahoo.com> wrote: