[BioC] fit=lmFit limma package

[Gordon K Smyth]

Dear Jing and Wolfgang,

lmFit() allows for NA/NaN/Inf values in the expression object, so simply 
having such values in the data isn't the problem.  However I agree with 
Wolfgang that there is likely to be a serious problem with the expression 
object to cause this error to occur.  I am guessing that the data is 
degenerate in some way.

Jing, we can't help you with this further given the limited information 
you've provided.  As Wolfgang says, you need to examine your data more 
closely.  A first step would be:

   summary(exprs(eset))

Best wishes
Gordon

> Date: Sun, 11 Sep 2011 23:17:22 +0200
> From: Wolfgang Huber <whuber at embl.de>
> To: bioconductor at r-project.org
> Subject: Re: [BioC] fit=lmFit limma package
>
> Jing
>
> have a look at the values
> 	 exprs(eset)
> Given the message you see, it appears that these values contain
> NA/NaN/Inf values. In that case, you need to investigate the procedure
> by which you created 'eset' for when & how that happens.
>
> 	Best wishes
> 	Wolfgang
>

> Sep/10/11 2:33 AM, Jing Huang scripsit::
>> Dear All members,
>>
>> I am trying to analyze  GEO data  and have used lots of ???limma??? package. Occasionally,  I got this R error :
>>
>>> library(limma)
>>> fit=lmFit(eset,design)
>> Error in lm.fit(design, t(M)) :
>>    NA/NaN/Inf in foreign function call (arg 4)
>>
>> Everything else seems right.
>>
>> Can somebody help me with this?
>>
>> Many Thanks
>>
>> Jing


______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}