Getting started, dummy's guide, weighted network chart

Jun 19, 2015 at 4:53 PM
I know that NodeXL is designed as a research tool for Social Media interaction.
I am wondering if I can use it for another purpose.
I'm just starting with network graphs, and want a simple example to start with.
I take a table of 20 towns and their distance apart.
I create a Vertex (Node) table of towns and population, and an Edge table of pairs of towns and label Edges!N2 as "Distance" and fill in the distances.
eg
Vertex 1 Vertex 2 Distance
Belfast Athlone 227
Cork Belfast 424
Cork Athlone 219

Now I want a chart.
I tried that and got a circular chart with equidistant points all connected which is not what I expected.
How do I draw a undirected graph which places these points in 2D space so that the length of the connector is proportional to the distance?

TIA
Patrick
Coordinator
Jun 19, 2015 at 7:43 PM
Hello!

Thank you for the interest in NodeXL.

NodeXL has several layout options. Most of these layouts are based on the idea that edge length is not fixed.

Usually edge strength is conveyed with edge width.

It is possible to fix the locations of the nodes so that the edge lengths do convey information. This would require that each vertex is assigned an x and y location. You can do this using the NodeXL>Visual Properties>Autofill Columns>Vertices>Vertex X location and Vertex Y location drop down lists. You would need to add two columns to the vertices worksheet, one for the X and one for the Y value. X and Y might be the latitude and longitude of each town.

Regards,

Marc
Jun 20, 2015 at 2:24 PM
Thanks, Marc.
So NodeXL is not what I should be using for this.
I started with towns because I figured I would know what the graph should roughly look like.

My real project will be to graph the similarity of text files of code, where the distance is the number of differences. Obviously there I would have no idea of, and do not care about, X & Y coordinates. What I need is an algorithm to group together any way in 2D the nodes with small distances, farther away from those a larger distance away.

I was advised to use Gephi but unfortunately their forums are closed to me as they use AHBL, a website which is now ceased operation, to give them a block list for IP addresses.

Best

Patrick