[BioC] Pairs plots in lumi, plots look different?

Kasper Daniel Hansen khansen at stat.Berkeley.EDU
Thu May 8 19:41:53 CEST 2008


Please don't put seed inside functions, it may mess up the random  
number stream. If someone wants reproducible plots you should either  
increase the number of points are let him set the seed himself.

Kasper

On May 8, 2008, at 7:35 AM, Pan Du wrote:

>
> Yes, as Matthias mentioned, we use random subset to increase the  
> efficiency
> of plotting. To avoid variations over different plots, I have added  
> the
> "seed" parameter to the these plot functions. Please check the latest
> developing version of lumi 1.7.3. Thanks for using lumi!
>
> Best regards,
>
>
> Pan
>
>
>
> On 5/8/08 5:00 AM, "bioconductor-request at stat.math.ethz.ch"
> <bioconductor-request at stat.math.ethz.ch> wrote:
>
>> Message: 1
>> Date: Wed, 07 May 2008 12:41:48 +0200
>> From: Matthias Kohl <Matthias.Kohl at stamats.de>
>> Subject: Re: [BioC] Pairs plots in lumi, plots look different?
>> To: Julien Bauer <jb393 at cam.ac.uk>
>> Cc: bioconductor at stat.math.ethz.ch
>> Message-ID: <4821876C.3020705 at stamats.de>
>> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>>
>> Hello,
>>
>> the default pairs plot uses a random subset of the data
>>
>> pairs(x, ..., logMode = TRUE, subset = 5000)
>>
>> confer
>> library(lumi)
>> ?"pairs-methods"
>>
>> Hence, each call of pairs leads to different results. By setting the
>> random seed or the argument "subset" appropriately you could obtain
>> identical plots for each call.
>>
>> Best regards,
>> Matthias
>>
>>
>> Julien Bauer wrote:
>>> Hello,
>>> I am working for a microarray facility at Cambridge University, we
>>> have been using the Illumina platform for a while now.
>>> Lumi is a great package but I noticed something rather odd, when  
>>> using
>>> the plot function "pairs" on my data, if I run it again the plots  
>>> look
>>> different, some points are shifted or change location. The rest stay
>>> the same it just the look of the graphs that change.
>>> My guess is that it is because of the auto scaling of the graph  
>>> but I
>>> would like to be sure. I look in the mailing list archive and in the
>>> vignette but I couldn't find the answer for this.
>>> Thanks in advance for your help,
>>>
>>> Julien Bauer
>>>
>>> _______________________________________________
>>> Bioconductor mailing list
>>> Bioconductor at stat.math.ethz.ch
>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>> Search the archives:
>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>
>> -- 
>> Dr. Matthias Kohl
>> www.stamats.de
>>
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor



More information about the Bioconductor mailing list