[R] What is the test statistics in perm.test

[Achim Zeileis]

On Fri, 23 Apr 2010, Greg Snow wrote:

> The perm.test function in exactRankTests makes reference to the StatXact 
> manual, so presumably that would have the exact details.  The details 
> section of the help file refers to converting real values to integers, 
> so I expect that it is not doing what many of us think of as a 
> permutation test, and I would not think that it is using t-statistics.

Well, it is similar to that and even more similar to the Wilcoxon 
statistic. But more importantly, perm.test() is superseded by 
independence_test() in package "coin" (to which the message at startup of 
"exactRankTests" points you), at least for two-sample tests.

The "coin" package is also much better documented, see in particular
   vignette("LegoCondInf", package = "coin")
which corresponds to the paper published in The American Statistician, see
   citation("coin")

> R is powerful enough that a permutation test can be run in just a few 
> lines of R code without needing a specialized function.  And when you do 
> it that way, you know exactly what is being done.

True, but in some situations there are more efficient sampling algorithms 
available, some of which are provided in "coin".

Best,
Z

> --
> Gregory (Greg) L. Snow Ph.D.
> Statistical Data Center
> Intermountain Healthcare
> greg.snow at imail.org
> 801.408.8111
>
> From: Xiang Gao [mailto:xianggao2006 at gmail.com]
> Sent: Friday, April 23, 2010 8:52 AM
> To: Greg Snow; djmuser at gmail.com
> Cc: r-help at r-project.org
> Subject: Re: [R] What is the test statistics in perm.test
>
> Thanks for all your reply.  I am a biologist and only start to use R recently. Yes, It is my fault. I assume that perm.test is a very popular method. Thanks for Dennis to point out that it is in exactRankTests package. I did not realize that there are so many R package. I guess I did not ask the right question and neither in the right way.
>
> Based on my limited knowledge on statistics, it seems like use t-statistics to test the permutation. I am just not sure.
>
> Xiang
>
>
>
>
>
> exactRankTests
> On Fri, Apr 23, 2010 at 9:13 AM, Greg Snow <Greg.Snow at imail.org<mailto:Greg.Snow at imail.org>> wrote:
> There is no perm.test function in base R, is this from a package? Which package? What does the documentation for that function/package say?
>
> --
> Gregory (Greg) L. Snow Ph.D.
> Statistical Data Center
> Intermountain Healthcare
> greg.snow at imail.org<mailto:greg.snow at imail.org>
> 801.408.8111
>
>> -----Original Message-----
>> From: r-help-bounces at r-project.org<mailto:r-help-bounces at r-project.org> [mailto:r-help-bounces at r-<mailto:r-help-bounces at r->
>> project.org<http://project.org>] On Behalf Of Xiang Gao
>> Sent: Thursday, April 22, 2010 1:43 PM
>> To: r-help at r-project.org<mailto:r-help at r-project.org>
>> Subject: [R] What is the test statistics in perm.test
>>
>> Hi,
>>
>> The permutation test for two samples in R is function perm.test(). I
>> could
>> not figure out what is the statistics it estimate and how many
>> permutation
>> it did in default?
>>
>> Thanks,
>>
>>       [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help at r-project.org<mailto:R-help at r-project.org> mailing list
>> 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.
>
>
>
> --
> Xiang Gao, Ph.D.
> Department of Biology
> University of North Texas
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> 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.
>