[BioC] DiffBind error loading dba.count

[Paolo Kunderfranco]

Hi Gordon,
I tried test=dba.count(test,bParallel=F) but still facing the same
problem, it seems that 6 Gb of RAM are not enough to load
2 GB of bed file, is it possibile?
I suggest I should switch to a bigger machine..
paolo



2012/7/31 Gordon Brown <Gordon.Brown at cancer.org.uk>:
> Hi, Paolo,
>
> DiffBind uses quite a lot of memory for the dba.count step, because it loads
> the whole data file into memory.  (In hindsight, not the best design
> choice...)  Try running with parallelization turned off:
>
>> test=dba.count(test,bParallel=F)
>
> so that it only loads one file at a time, to reduce the memory footprint.
> For the upcoming release, we'll probably fix this, at the cost of somewhat
> slower counting.
>
> How much RAM does your machine have?
>
> Cheers,
>
>  - Gord
>
>
> On 2012/07/30 11:00, "Paolo Kunderfranco" <paolo.kunderfranco at gmail.com>
> wrote:
>
>> Dear All,
>> I am facing some problems with RGUI loading dba.count with my dataset,
>>
>>> setwd("C:/Documents and Settings/pkunderfranco/Documenti/Data/ChIP-seq
>>> Roberto/Analisi_PK/DiffBind/H3K27me3/")
>>
>>> library("DiffBind")
>>
>>> test= dba(sampleSheet='database.csv')
>> ES_H3K27 ES H3K27 1  1 raw
>> CMN_H3K27 CMN H3K27 1  1 raw
>> CMA_H3K27 CMA H3K27 1  1 raw
>>
>>> test=dba.count(test)
>> AppName: rgui.exe  AppVer: 2.150.58871.0  ModName: diffbind.dll
>> ModVer: 0.0.0.0  Offset: 00003be1
>>
>> Did anyone faced the same problem? which are the memory requirements
>> to run DiffBind?
>> Thanks,
>> Paolo
>>
>>
>
>
> NOTICE AND DISCLAIMER
> This e-mail (including any attachments) is intended fo...{{dropped:16}}