[BioC] memory problem with fitPLM in package affyPLM

[fhong@salk.edu]

Thanks, Ben. That helps a lot! But I still have some questions? Would you
please also help me on this.

> There were significant changes in the structure of the PLMset object
> between 1.2.x and 1.3.x which is why you are having problems with the
> boxplot(), Mbox() commands on your old PLMset using the new code.
But why when I reload in to R 2.0.1 ( the on ei used to generate PLMset
object), and tried boxplot ( suppose to produce NUSE plot), it gave me
something strange (see attachment)


>
> To reduce memory usage with fitPLM I suggest that you do something like
> the following
>
> Pset <- fitPLM(my.Data,output.param=list(varcov="none"))
>
> which should significantly reduce memory usage (and this should really
> be the default setting, an oversight on my part which will be fixed
> shortly).
>
>
> Also, if you can live without the weights (or alternatively the
> residuals) you could do
>
> Pset <- fitPLM(my.Data,output.param=list(varcov="none",weights=FALSE))
>
>
> or
>
> Pset <- fitPLM(my.Data,output.param=list(varcov="none",residuals=FALSE))
>
> which would also reduce the memory overhead.
Will those simplificaiton change the underlying model that is fitted to
the data. e.g., weights=FALSE doesn't this mean it won't use iteratively
reweighted least squares (IRLS)?

Thanks!

Fangxin


> On Thu, 2005-06-09 at 10:22 -0700, fhong at salk.edu wrote:
>> Sorry to bring such an old question up again. I hope that somebody can
>> explain what is happening here.
>>
>> (1)
>> I tried to run fitPLM to about 72 affy ATH1 arrays. It worked with
>> affyPLM
>> version 1.2.5 under R2.0.1, then I saved it as "dataPLM.RData" with
>> option
>> "compress=TRUE"
>> Later when I reload "dataPLM.RData" into working space (R2.1.0) and try
>> to
>> run boxplot and Mbox, it told me that
>>
>> " Error in if (object at model.description$R.model$which.parameter.types[3]
>> ==
>> :
>>         argument is of length zero
>>  since object at model.description is NULL."
>>
>> How did this happen? Does "compress=TRUE" destroy some of the internal
>> structure of PLMset object?
>>
>> (2)
>> then I tried to run fitPLM again ( on the affyBatch object I saved
>> before)
>> using package affyPLM version 1.3.3 under R2.1.0 But this time it
>> wouldn't
>> finish (after Background correcting PM, Normalizing PM)
>>  but generated an error of R and closed R after that. I thought that may
>> because my laptop has only 1.5Gb memory.
>> Does this mean that there is some substantial change of affyPLM package
>> between version 1.3.3 and version 1.2.5?
>>
>>
>> I then tried to run this using a Linux machine with 4Gb accessible
>> memory
>> size (there is no limitation in memory for each user and I reboot the
>> system as a super user). But I still got the error message about memory,
>> saying
>>
>> > Pset.Goda=fitPLM(data.Goda.I)
>> Background correcting PM
>> Normalizing PM
>> Fitting models
>> Error: cannot allocate vector of size 52 Kb
>>
>> Does this mean 4Gb is still not enough?
>>
>> My last question is what I need to update my computer in order to
>> analyze
>> 72 arrays together (simpleaffy, affyPLM,Limma...)
>>
>>
>> Thank you very much, any suggestion will be highly appreciated!
>> Fangxin
>>
>>
>>
>>
>> --------------------
>> Fangxin Hong  Ph.D.
>> Plant Biology Laboratory
>> The Salk Institute
>> 10010 N. Torrey Pines Rd.
>> La Jolla, CA 92037
>> E-mail: fhong at salk.edu
>> (Phone): 858-453-4100 ext 1105
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at stat.math.ethz.ch
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
> --
> Ben Bolstad <bolstad at stat.berkeley.edu>
> http://www.stat.berkeley.edu/~bolstad
>
>
>


--------------------
Fangxin Hong  Ph.D.
Plant Biology Laboratory
The Salk Institute
10010 N. Torrey Pines Rd.
La Jolla, CA 92037
E-mail: fhong at salk.edu
(Phone): 858-453-4100 ext 1105