[Rsiena-help] Converting to Siena objects--memory overload

[Ruth Ripley]

Dear Jennifer,

I began to run this and it slowed down drastically as soon as I started 
to use the network package. You use an array of matrices for sienaNet 
not networks. To save time and space I would avoid doing anything from 
the network and sna packages in the same run as RSiena. RSiena is 
completely independent of them apart from using read.paj to convert 
pajek data files to matrices.

I would try running RSiena on your own machine without the network 
things. You can use sparse version of the matrices to save space if you 
wish. (If you can't find the code for the conversion in the manual, 
please ask for help).

Regards

Ruth

On 17/02/2012 14:39, Victor, Jennifer Nicoll wrote:
> Hello Ruth:
>
> Thank you VERY much for offering to help with this.  I'm in quite a bind and could use some advice.  I am attaching my R script and my data (not for distribution, please!) so that you can help me determine where I'm going wrong.  I really appreciate your advice!!
>
> Best,
> Jennifer
>
>
> Jennifer N. Victor
> Assistant Professor of Political Science
> University of Pittsburgh
> 4600 Posvar Hall
> jnvictor at pitt.edu
> (412) 624-7204
>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:<http://lists.r-forge.r-project.org/pipermail/rsiena-help/attachments/20120216/11d57fa8/attachment-0001.html>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 16 Feb 2012 18:50:25 +0000
> From: Ruth Ripley<ruth at stats.ox.ac.uk>
> To: rsiena-help at lists.r-forge.r-project.org
> Subject: Re: [Rsiena-help] Converting to Siena objects--memory
>          overload
> Message-ID:<4F3D4FF1.7050008 at stats.ox.ac.uk>
> Content-Type: text/plain; charset=windows-1252; format=flowed
>
> Dear Jennifer,
>
> I don't understand this, as sienaNet does very little processing and
> your objects should not be huge. I suggest you stop the process.
>
> Please can you help me with a bit more information: do the command
> sessionInfo() within an R session and send the result.
>
> It might also help to see the R commands you used to create the
> matrices. And if you would like to send me the data privately I could
> try it out for you.
>
> Regards,
>
> Ruth
>
> On 16/02/2012 18:14, Victor, Jennifer Nicoll wrote:
>> Hello users and experts:
>>
>> I am attempting to create a Siena object with my 3-wave matrices and
>> it?s taking an enormous amount of memory. I have three matrices of size
>> 555x555 and I have attempted the following command to create Siena
>> objects for my dependent variables:
>>
>> caucuses<-sienaNet(array(c(cauc109.net, cauc110.net, cauc111.net),
>> dim=c(555, 555, 3)))
>>
>> I have been running this on a virtual ?supercomputer?: Dell PowerEdge
>> with Dual Intel X5650 Six Core Processors and 24 GM Memory. The datafile
>> is 10MB and the command has been running for 15 hours now (so far no
>> errors).
>>
>> Shall I continue to let it run? Have I done something wrong? I shudder
>> to think what will happen when I try to estimate a model!
>>
>> Thanks for your advice,
>>
>> Jennifer Victor
>>
>> ___________________________________________
>> Jennifer Nicoll Victor
>> Assistant Professor
>> Department of Political Science
>> University of Pittsburgh
>> 4600 Wesley W. Posvar Hall
>> (412) 624-7204
>> E-mail: jnvictor at pitt.edu<mailto:jnvictor at pitt.edu>
>> Homepage: http://www.polisci.pitt.edu/person/jennifer-nicoll-victor
>>
>>
>>
>> _______________________________________________
>> Rsiena-help mailing list
>> Rsiena-help at lists.r-forge.r-project.org
>> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rsiena-help

-- 
Ruth M. Ripley,                         Email:ruth at stats.ox.ac.uk
Dept. of Statistics,                    http://www.stats.ox.ac.uk/~ruth/
University of Oxford,                   Tel:   01865 282857
1 South Parks Road, Oxford OX1 3TG, UK  Fax:   01865 272595