How do I convert a bimodal network?

Jul 11, 2011 at 4:34 AM


I have a network of people and the projects they worked on, for example: fred, sally and david all worked on project alpha; sally and david worked on project beta; and fred worked on project gamma by himself. I think its called a bimodal network.

If two or more people collaborated on a project I'd like to consider them "connected", and visualise the connections directly between people (i.e. not see connections via project nodes on the graph). That way I can see who is the most collaborative, who works in isolation etc.

The network is 706 rows long, so I need an automated way to do the conversion. Can anyone help?


Jul 11, 2011 at 5:19 PM
Edited Jul 11, 2011 at 6:10 PM

I don't know of any way to do that directly with Excel features or formulas.  I think it requires some programming, although the program would be pretty simple.  It could be done inside an Excel workbook using VBA, Office's built-in language, or in any other language that can parse and process text.

Perhaps someone else has a better idea.

-- Tony

Jul 13, 2011 at 3:22 AM

Thanks Tony. If anyone has any further ideas, or can contribute the relevant code for the VBA program I'd really appreciate it.

Aug 8, 2011 at 1:28 AM

I ended up using a program called UCINet to convert the two-mode data into two "one-mode" data sets. UCINet has a built-in tool called "Affiliations" for just this scenario. You can read more about it here:

Aug 8, 2011 at 11:06 PM

Thanks for letting us know.

-- Tony