[R] candisc() error message

[Michael Friendly]

John is correct & the error message you got was clear:  the error SSP 
matrix is of deficient rank.
Possible solutions:
- Regroup the factor levels to fewer and/or
- Combine some of the species to broader classes so you have fewer than 
17 response variables
- Collect more samples

hope this helps,
-Michael

John Fox wrote:
> Dear Tanya,
>
> If I follow correctly what you're trying to do, you're in effect doing a
> one-way MANOVA of 17 response variables on a six-level factor, with only 20
> observations. That's not enough data. I'm copying to Michael Friendly in
> case he hasn't seen your posting and would like to comment.
>
> Regards,
>  John
>
> ------------------------------
> John Fox, Professor
> Department of Sociology
> McMaster University
> Hamilton, Ontario, Canada
> web: socserv.mcmaster.ca/jfox
>
>
>   
>> -----Original Message-----
>> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
>>     
> On
>   
>> Behalf Of Tanya Yatsunenko
>> Sent: June-15-08 8:23 PM
>> To: r-help at r-project.org
>> Subject: [R] candisc() error message
>>
>> Hi,
>> I am doing canonical discriminant analysis using candisc function from
>> the candisc package.
>> My input is a table of species distribution (columns = abundance of each
>> species in each sample)  in samples that are split by categories (rows),
>> and I want to know whether each category is associated with a particular
>> set of species and their abundances.
>> I have 20 rows (samples) split into 6 categories, and 17 columns
>>     
> (species).
>   
>> I am getting the following error message, which I don't understand:
>>
>>  > can<-candisc(mod, data=canIN)
>> Error in linear.hypothesis.mlm(mod, hyp.matrix.2, SSPE = SSPE, V = V,  :
>>   The error SSP matrix is apparently of deficient rank = 16 < 17
>>
>> Does anyone has any experience in candisc?
>>
>> --
>> Tanya.
>>
>>
>> 	[[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.
>>     
>
>   


-- 
Michael Friendly     Email: friendly at yorku.ca 
Professor, Psychology Dept.
York University      Voice: 416 736-5115 x66249 Fax: 416 736-5814
4700 Keele Street    http://www.math.yorku.ca/SCS/friendly.html
Toronto, ONT  M3J 1P3 CANADA