[Rsiena-help] Matching matrices

[Matthew Weber]

Thanks Nathan ­ especially for the quick response!

The problem is that the matrices don¹t match up in any way. The actors are
the same from matrix 1 to matrix 2 to matrix 3, but if actor D is not
present at time 2, he simply doesn¹t show up in the matrix.

In other words, if matrix 1 has the following actors:

1, 2, 3, 4, 5, 6, 7, 8, 9, 10

But actors 3, 6, 9 and 10 are not present in the matrix at time 2, then the
header for matrix 2 would be as follows:

1, 2, 4, 5, 7, 8

So I would then need to reinsert the missing headers and the structural
zeros.

I could be missing something from what you¹ve said, but I¹ve been through
the manual a few times looking for solutions to this and wasn¹t able to find
a clear answer. I¹d like to find a way to code this through R, or even to
force a shortcut in UCINET, because with 8 waves and up to 2000 actors in
each wave.

As I said, I may be missing something...

Thanks Again,
Matt

On 6/24/11 3:30 PM, "Nate Doogan" <doogan.1 at osu.edu> wrote:

> Hey Matt.
> 
> Take a look at the RSiena manual for some examples. There's some examples with
> Pajek data, specifically.
> 
> Generally speaking, I believe there are two methods of indicating structurally
> missing values--structural zeros (value 10) placed directly within the
> matrices, and the creation of an exogenous events file (manual section 2.1).
> The latter allows more precision if you know exactly when actors left the
> setting.
> 
> Nathan Doogan
> Doctoral Student
> The Ohio State University
> 
> On Fri, Jun 24, 2011 at 3:21 PM, Matthew Weber <mattsweber at gmail.com> wrote:
>> Hi All,
>> 
>> I'm working with a new set of longitudinal data tracking interactions
>> between members of a large corporation. I have eight waves of data in a
>> network format, but the number of actors change by wave - anywhere from 1200
>> to 2000 + basic attributes.
>> 
>> I'd like to analyze the behavioral patterns in this data using RSiena, but
>> am not quite sure how to handle the changes in the number of actors.
>> 
>> Is there an easy way in R to insert structural zeros for this data? Any
>> suggestions on the best format to use?
>> 
>> I currently have the data in matrix form and pajek.net <http://pajek.net> .
>> I'm stumped as to
>> how best to get this into a format that RSiena will work with, and my brain
>> is a bit fried. Any help is greatly appreciated.
>> 
>> Thanks,
>> Matt
>> 
>> 
>> 
>> ==================================
>> 
>> Matthew S. Weber
>> Postdoctoral Research Associate
>> Center for Technology, Entertainment and Media (CTEM)
>> Fuqua School of Business
>> Duke University
>> 
>> matthew.weber at duke.edu
>> (ph) 919.660.2894 <tel:919.660.2894>
>> 
>> ==================================
>> 
>> 
>> 
>> _______________________________________________
>> 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
> 
> 

==================================

Matthew S. Weber
Postdoctoral Research Associate
Center for Technology, Entertainment and Media (CTEM)
Fuqua School of Business
Duke University

matthew.weber at duke.edu
(ph) 919.660.2894

==================================


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rsiena-help/attachments/20110624/2a47d80e/attachment.htm>