GLM terminating - too many predictors?

[Lambodhar Damodaran]

Hi everyone,

I am having an issue where I get this error when trying to run a GLM :

Adding independent predictors 'Region.coefficients' with design matrix 'Region.designMatrix'
    Total # of predictors = 34 and rank = 7
    Provided matrix of independent variables is not identifiable.

dr.app.beast.BeastMain <init>
SEVERE: Parsing error - poorly formed BEAST file, regdiv-2012-2013.700.GLM.xml:
Error parsing '<glmModel>' element with id, 'null':
'id' attribute was not found in glmModel element.

Error thrown at: dr.xml.XMLObject.getAndTest(Unknown Source)

When I run it with fewer predictors (~6) it will run without issues but it doesn't seem to do well with more than that. Can someone help me figure out what's going here? Also, what does the rank number mean in this error message? Any help would be greatly appreciated!

Thanks,

Lambodhar

[Jonathon Gass]

Dear all, reasonably urgent request...

I am experiencing the exact same error as lambo above. Does anyone have insight into what this error is about? I have 27 predictors and rank = 18. I have not tried yet to reduce the predictors as I would prefer to include all 27. 

Thanks for your help!

JDG