***** To join INSNA, visit http://www.insna.org *****
Hi Amoshaun,
Although in development, we at CNS have developed a little program - Ceunet - for data handling (and visualization and very rudimentary analysis) that contains a function for merging actors (menu Matrix operations/Merge actors). It can import/export to several data formats (Pajek, DL, csv etc) and handles multiple matrices (and attributes) so it might do the trick for you. However, although it works fine for sociomatrices when merging actors, any attributes (string, double or boolean) are discarded in the process - need to fix that... That is: still in development and primarily for internal use, and occassionally buggy (for instance, if you got the actor tab open before merging, you need to close and reopen it to see the new actor set...)
Ceunet runs on Windows - download here (~500 kB):
http://cnslabs.ceu.hu/
(Also needs .NET Framework 4.0 installed, which is a standard Windows package - if needed, it should download that automatically from microsoft)
Yours,
Carl
Carl Nordlund, PhD
Post-doctoral research fellow
Center for Network Science, Department of Political Science
Central European University, Budapest
>>> atoft 12-08-21 20:51 >>>
***** To join INSNA, visit http://www.insna.org *****
Hi all!
We are working on a project on anti-human trafficking networks and we
are looking for a script for merging multiple nodes in a relational
network matrix. I have done it manually by summing the edges for a given
set of nodes and renaming as a summed node (i.e. counting all members of
a family as a single family node rather than individual nodes). This is
too cumbersome for multiple datasets, so we are looking for a function
that merges nodes based on a script to produce a transformed dataset,
preferably in R or UCINet, but Pajek or something else would be OK too.
We would then use this transformed dataset to run network metrics like
in-degree/out-degree, etc. at the Producer level.
Some context: We are analyzing issuecrawler.net hyperlink network files
that read as directed out-link/in-link counts based on a URL to URL
matrix. However, most of our coded data is read at what we call the
Producer level. The problem is that some Producers are responsible for
multiple URLs. To analyze the hyperlink matrix alongside the coded data
at the Producer level, we need to merge a given set of URL nodes into a
single Producer node.
Thank you for any suggestions you might have!
Amoshaun
--
Amoshaun Toft, PhD
Lecturer
Interdisciplinary Arts and Sciences
University of Washington, Bothell
UW 1-143 ; 425-352-3210
[log in to unmask] ; @amoshaun
http://atoft.net/
_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.
_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.
|