[Rsiena-help] Transform 1-mode edgelist to adjacency matrix (Daniel Stefan Hain)

[Christoph Stich]

Dear Daniel,

The SIENA Manual actually provides instructions of who to deal with the 
problem of leaving and joining actors. The relevant sections are "4.1.2 
Structurally determined values" and "4.8 Composition change: joiners and 
leavers". In general, you either specify an actor as absent from the 
network by setting the value of his ties as "structural zeros", you set 
ties of absent actors to missing or you specify absent actors with 
sienaCompositionChangeFromFile or sienaCompositionChange.

If I understand you correctly you transform your adjacency matrix into 
an igraph object, subset for your desired attributes, then back into an 
adjacency matrix and on the way you lose some of your actors? You could 
insert rows and columns by hand (insertCol and insertRow from the 
micEcon package do that), but that seems rather cumbersome. Have you 
tried subsetting your ties directly from the adjacency matrix? Maybe 
transforming your year attribute in an attribute matrix and then using 
that to subset your original matrix?


Hope that helps,
Christoph Stich
MSc candidate Urban Studies, University of Amsterdam



On Fri, Nov 30, 2012 at 3:00 AM, 
<rsiena-help-request at lists.r-forge.r-project.org 
<mailto:rsiena-help-request at lists.r-forge.r-project.org>> wrote:

    Send Rsiena-help mailing list submissions to
    rsiena-help at lists.r-forge.r-project.org
    <mailto:rsiena-help at lists.r-forge.r-project.org>

    To subscribe or unsubscribe via the World Wide Web, visit
    https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rsiena-help

    or, via email, send a message with subject or body 'help' to
    rsiena-help-request at lists.r-forge.r-project.org
    <mailto:rsiena-help-request at lists.r-forge.r-project.org>

    You can reach the person managing the list at
    rsiena-help-owner at lists.r-forge.r-project.org
    <mailto:rsiena-help-owner at lists.r-forge.r-project.org>

    When replying, please edit your Subject line so it is more specific
    than "Re: Contents of Rsiena-help digest..."


    Today's Topics:

        1. Transform 1-mode edgelist to adjacency matrix (Daniel Stefan
    Hain)


    ----------------------------------------------------------------------

    Message: 1
    Date: Thu, 29 Nov 2012 15:24:03 +0000
    From: Daniel Stefan Hain <dsh at business.aau.dk
    <mailto:dsh at business.aau.dk>>
    To: "rsiena-help at lists.r-forge.r-project.org
    <mailto:rsiena-help at lists.r-forge.r-project.org>"
             <rsiena-help at lists.r-forge.r-project.org
    <mailto:rsiena-help at lists.r-forge.r-project.org>>
    Subject: [Rsiena-help] Transform 1-mode edgelist to adjacency matrix
    Message-ID:
            
    <5C7C94834D71E144B9351830389C46BCFCDBFFC1 at AD-EXCHMBX2-4.aau.dk
    <mailto:5C7C94834D71E144B9351830389C46BCFCDBFFC1 at AD-EXCHMBX2-4.aau.dk>>
    Content-Type: text/plain; charset="iso-8859-1"

    Hi dear RSiena community,

    I currently have the following problem. I am conducting a dynamic
    analysis on cooperation in research project on basis of a directed
    edgelist that looks like follows:

    "Project leader"               "Project partner" "Year"            
           "Some more attributes"
    Firm X                                   Firm Y                    
      2000
    Firm A                                   Firm X                     2005
    ...

    I want to divide the dataset into 3-4 subsets for different periods,
    lets say 2000-2002, 2003- 2005 ect. That's not the challenge
    (subseting in [R]), but then my problem is the following:

    1.: How can I create a directed and weighted adjacency matrix out of
    this data structure for every single time period?  Up to now I
    managed it with transforming it to an IGraph object and then back
    into an adjacency matrix. That works,  but the resulting matrix for
    the selected timeframe does not contain all actors from the whole
    dataset, thus the resulting adjacency matrixes  as input for RSiena
    all have different dimensions.

    I up to now worked with 2-mode edgelists, there I just could
    tabulate the dataframe and multiply it with the transposed one to
    get the adjacency matrix. When I used subsets for the different time
    period, I could exploit the fact that R still keeps the index of
    everything that drops out of the subset, but shows up in the
    cross-tab and therefore in the adjacency again.

    With a 1-mode edgelist it obviously does not work. Any advice is
    appreciated!

    Vest wishes

    Daniel S. Hain
    PhD Fellow, M.Sc. Economics, Dipl.-Wirt.-Ing.
    _______________________________________
    Innovation, Knowledge and Economic Dynamics (IKE) Group
    Strategic Research Alliance for Energy Innovation Systems (EIS)
    Department of Business and Management
    Aalborg University
    Fibigerstraede 4, Room 16a
    9220 Aalborg ?st, Denmark

    Tel.:    +45 9940 2724
    Email: dsh at business.aau.dk
    <mailto:dsh at business.aau.dk><mailto:dsh at business.aau.dk
    <mailto:dsh at business.aau.dk>>
    Web: www.business.aau.dk
    <http://www.business.aau.dk><http://www.business.aau.dk>

    -------------- next part --------------
    An HTML attachment was scrubbed...
    URL:
    <http://lists.r-forge.r-project.org/pipermail/rsiena-help/attachments/20121129/ba8c60bd/attachment-0001.html>

    ------------------------------

    _______________________________________________
    Rsiena-help mailing list
    Rsiena-help at lists.r-forge.r-project.org
    <mailto:Rsiena-help at lists.r-forge.r-project.org>
    https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rsiena-help

    End of Rsiena-help Digest, Vol 28, Issue 1
    ******************************************




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rsiena-help/attachments/20121130/a708fab5/attachment.html>