[R] Help with perm.test please

About this list Date view Thread view Subject view Author view Attachment view

From: - - (mbatobeme@yahoo.com)
Date: Thu 05 Sep 2002 - 14:42:15 EST 

Message-id: <20020905044215.54927.qmail@web10506.mail.yahoo.com>

Hi All:

OK, this is probably a stupid thing to do but.......

Say I have 706 controls and 6 treatments, and want to
do a basic two-sample permutation test. I could do a
monte carlo approximation but instead want to use
perm.test.

Now it happens that my data is counts with
> unique(x)
[1] 1 2
> unique(y)
 [1] 1 2 5 3 10 9 7 4 6 11 8 23 18 30 16 29
42

so there aren't that many categories.

I thought perm.test makes this an Rx2 table and then
does some sort of network algorithm on it.

But I get the following error message:

> perm.test(x,y,exact=T)
Error in cperm(eq, m, paired) : N > 200 in cpermdistr2
>

I understand this is a sample size error but it
doesn't seem to make sense to me since there aren't
that many unique categories for x and y. What am I
missing ?

Thanks,

Steve

__________________________________________________

Yahoo! Finance - Get real-time stock quotes

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

About this list Date view Thread view Subject view Author view Attachment view

This archive was generated by hypermail 2.1.3 : Tue 01 Jun 2004 - 09:31:12 EST