Motif SImplification using .net libraries and Large data analysis

Sep 4, 2013 at 10:20 AM
Edited Sep 4, 2013 at 11:25 AM
Hi,

I have two questions -

1). Can we use motif simplifications in .net applications using the nodexl libraries. I can draw graphs easily but would like to know how to simplify clusters using motif simplifications? A little guidance is appreciated.

2). Has the community considered using layout methods like 'OpenOrd as used in by Gephi' for large network analysis (around 200000+ nodes).

Thank You.

Regards,
Abhi
Sep 5, 2013 at 6:10 PM
Abhi:

The low-level code for detecting motifs is exposed in the NodeXL Class Libraries--see the Smrf.NodeXL.Algorithms.MotifCalculator class. However, the higher-level code that handles the visualization of motifs, which is what you're interested in, is not exposed in the Libraries. That turned out to be difficult to generalize, and we ended up implementing it in our application-level code instead. The result is that we don't support motif simplifications in other applications.

If you're feeling ambitious, you can download the source code and look at what our application does to show collapsible motifs, then adapt the code to your own application. A starting point is the Smrf.NodeXL.ExcelTemplate.MotifCalculator2 and Smrf.NodeXL.ExcelTemplate.CollapsedGroupAttributeAdder classes. It's not rocket science, but it's not as easy as I'm sure you would like it to be.

Thanks for the pointer on the layout algorithm. I'm going to take a look at it.

-- Tony