<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7654.12">
<TITLE>RE: [Rsiena-help] Interaction effect specification for bipartitenetworks</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Dear Ruth,<BR>
<BR>
I had the same problem, I use 8 processors. Is it fixed already?<BR>
<BR>
Best regards,<BR>
<BR>
Chip<BR>
<BR>
-----Oorspronkelijk bericht-----<BR>
Van: rsiena-help-bounces@lists.r-forge.r-project.org namens marco tonellato<BR>
Verzonden: wo 17-3-2010 15:38<BR>
Aan: Ruth M. Ripley<BR>
CC: rsiena-help@r-forge.wu-wien.ac.at<BR>
Onderwerp: Re: [Rsiena-help] Interaction effect specification for bipartitenetworks<BR>
<BR>
Dear Ruth,<BR>
<BR>
That was quick! Meanwhile, I've tried running the model with only one core<BR>
and it worked as you predicted. I'll try the new code as soon as it'll be<BR>
available. Thanks a lot for your help!<BR>
<BR>
Best,<BR>
<BR>
Marco<BR>
<BR>
<BR>
On Wed, Mar 17, 2010 at 2:26 PM, Ruth M. Ripley <ruth@stats.ox.ac.uk> wrote:<BR>
<BR>
> Dear Marco,<BR>
><BR>
> I have now put the new source on R-forge. I hope it will be built<BR>
> by tomorrow.<BR>
><BR>
> Regards,<BR>
><BR>
> Ruth<BR>
> --<BR>
> Ruth M. Ripley, Email:ruth@stats.ox.ac.uk<Email%3Aruth@stats.ox.ac.uk><BR>
> Dept. of Statistics, <A HREF="http://www.stats.ox.ac.uk/~ruth/">http://www.stats.ox.ac.uk/~ruth/</A><<A HREF="http://www.stats.ox.ac.uk/%7Eruth/">http://www.stats.ox.ac.uk/%7Eruth/</A>><BR>
> University of Oxford, Tel: 01865 282851<BR>
> 1 South Parks Road, Oxford OX1 3TG, UK Fax: 01865 272595<BR>
><BR>
><BR>
> On Wed, 17 Mar 2010, Ruth M. Ripley wrote:<BR>
><BR>
> > Dear Marco,<BR>
> ><BR>
> > I have tried a similar model myself which fails... I know what is wrong:<BR>
> > user specified interactions will not work with multiple processes. It<BR>
> will<BR>
> > take me a little while to fix it, and it is late here. I will try<BR>
> > tomorrow!<BR>
> ><BR>
> > Regards,<BR>
> ><BR>
> > Ruth<BR>
> > --<BR>
> > Ruth M. Ripley, Email:ruth@stats.ox.ac.uk<Email%3Aruth@stats.ox.ac.uk><BR>
> > Dept. of Statistics, <A HREF="http://www.stats.ox.ac.uk/~ruth/">http://www.stats.ox.ac.uk/~ruth/</A><<A HREF="http://www.stats.ox.ac.uk/%7Eruth/">http://www.stats.ox.ac.uk/%7Eruth/</A>><BR>
> > University of Oxford, Tel: 01865 282851<BR>
> > 1 South Parks Road, Oxford OX1 3TG, UK Fax: 01865 272595<BR>
> ><BR>
> ><BR>
> > On Tue, 16 Mar 2010, marco tonellato wrote:<BR>
> ><BR>
> > > Dear Ruth,<BR>
> > ><BR>
> > > now that I changed the specification as 'ego' effect I could start the<BR>
> > > model, but R then crashed.<BR>
> > ><BR>
> > > This is the command I used and the output I got after the crash:<BR>
> > ><BR>
> > > ><BR>
> > ><BR>
> myeff<-includeInteraction(myeff,egoX,outActSqrt,interaction1="dev.ten",include=TRUE)<BR>
> > > name shortName type<BR>
> interaction1<BR>
> > > collaboration.objEffects.138.1 collaboration unspInt eval<BR>
> > > interaction2 include effect1 effect2<BR>
> effect3<BR>
> > > collaboration.objEffects.138.1 TRUE 24 17<BR>
> 0<BR>
> > ><BR>
> > > > ans <- siena07(mymodel, data=mydata, effects=myeff, batch=TRUE,<BR>
> > > verbose=TRUE, nbrNodes=3, useCluster=TRUE, initC=TRUE)<BR>
> > ><BR>
> > > Stochastic approximation algorithm.<BR>
> > > Initial value for gain parameter = 0.3464102.<BR>
> > > Start of the algorithm.<BR>
> > > Target function values are<BR>
> > > 1. 904.00000 2. 89.00000 3. 1009.38249 4. 5637.09646 5.<BR>
> > > -23.27556<BR>
> > > 6. 354.12048 7. 72.53726 8. 221.19411 9. -16.00520 10.<BR>
> > > 2756.24970<BR>
> > > 11. 2756.24970<BR>
> > ><BR>
> > > Start phase 0<BR>
> > > theta: -1.55 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00<BR>
> 0.00<BR>
> > > Current parameter values:<BR>
> > > -1.547494 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000<BR>
> > > 0.000000<BR>
> > > 0.000000 0.000000 0.000000<BR>
> > ><BR>
> > > Start phase 1<BR>
> > > Error in unserialize(node$con) : error reading from connection<BR>
> > ><BR>
> > ><BR>
> > > Best regards,<BR>
> > ><BR>
> > > Marco<BR>
> > ><BR>
> > > On Tue, Mar 16, 2010 at 11:28 PM, Ruth M. Ripley <ruth@stats.ox.ac.uk<BR>
> >wrote:<BR>
> > ><BR>
> > > > Dear Marco,<BR>
> > > ><BR>
> > > > Sorry, I think the information I have coded in the package is<BR>
> incorrect. I<BR>
> > > > think the egoX should be an 'ego' effect not a 'dyadic': if you<BR>
> change<BR>
> > > > this in your effects object it should be OK.<BR>
> > > ><BR>
> > > > I will fix it in the package but it will be a couple of days before<BR>
> it is<BR>
> > > > available in the binary on R-forge.<BR>
> > > ><BR>
> > > > Regards,<BR>
> > > ><BR>
> > > > Ruth<BR>
> > > > --<BR>
> > > > Ruth M. Ripley, Email:ruth@stats.ox.ac.uk<Email%3Aruth@stats.ox.ac.uk><BR>
> <Email%3Aruth@stats.ox.ac.uk <Email%253Aruth@stats.ox.ac.uk>><BR>
> > > > Dept. of Statistics,<BR>
> <A HREF="http://www.stats.ox.ac.uk/~ruth/">http://www.stats.ox.ac.uk/~ruth/</A> <<A HREF="http://www.stats.ox.ac.uk/%7Eruth/">http://www.stats.ox.ac.uk/%7Eruth/</A>><<BR>
> <A HREF="http://www.stats.ox.ac.uk/%7Eruth/">http://www.stats.ox.ac.uk/%7Eruth/</A>><BR>
> > > > University of Oxford, Tel: 01865 282851<BR>
> > > > 1 South Parks Road, Oxford OX1 3TG, UK Fax: 01865 272595<BR>
> > > ><BR>
> > > ><BR>
> > > > On Tue, 16 Mar 2010, marco tonellato wrote:<BR>
> > > ><BR>
> > > > > Dear All,<BR>
> > > > ><BR>
> > > > > I'm currently modeling a bipartite network of 83 x 577 nodes with<BR>
> RSiena<BR>
> > > > > (revision 0.61). I was able to estimate significant and convergent<BR>
> > > > > parameters for the endogenous part and the main effects of 5<BR>
> attributes<BR>
> > > > > (related to either node set).<BR>
> > > > ><BR>
> > > > > I now have a problem including interaction effects. In particular I<BR>
> would<BR>
> > > > > like to include the interaction of a constant binary covariate<BR>
> (which<BR>
> > > > > captures whether or not my nodes of the first mode have tenure)<BR>
> with the<BR>
> > > > > outdegree activity.<BR>
> > > > ><BR>
> > > > > Both using the tcl/tk initerface and entering (also with RSienaTest<BR>
> > r62):<BR>
> > > > ><BR>
> > > > ><BR>
> > > ><BR>
> myeff<-includeInteraction(myeff,egoX,outActSqrt,interaction1="dev.ten",include=TRUE)<BR>
> > > > ><BR>
> > > > > R returned this error message:<BR>
> > > > ><BR>
> > > > > invalid interaction specification:"must be at least one ego or both<BR>
> > > > dyadic<BR>
> > > > > effects".<BR>
> > > > ><BR>
> > > > > Finally, I also noticed that the "interactionType" attribute of<BR>
> the<BR>
> > > > tenure<BR>
> > > > > egoX effect is set as dyadic, whereas it is left unspecified for<BR>
> all the<BR>
> > > > > altX effects related to the other attributes I measured for the<BR>
> nodes of<BR>
> > > > the<BR>
> > > > > second mode.<BR>
> > > > ><BR>
> > > > > Am I doing something wrong? What should I do to get the interaction<BR>
> > > > effects<BR>
> > > > > working?<BR>
> > > > > Thanks a lot,<BR>
> > > > ><BR>
> > > > > Marco<BR>
> > > > ><BR>
> > > ><BR>
> > > ><BR>
> > ><BR>
> ><BR>
> ><BR>
> ><BR>
><BR>
><BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>