1

I have a graph showing nodes on the map of the US. There are between 100 and 400 nodes, and all two nodes have a line connecting them.

The problem is that there are so many connections that the graph becomes virtually meaningless. To deal with this, I have removed, through some simple algorithm, about 10% of nodes each time. Still, however, the graph is not very readable. This is what it looks like with 10% removed:

Sample

The purpose of the graph is to see which connections are strongest (the brightest lines indicate strongest connections). But, as I'm sure you would agree, it's pretty hard to interpret anything.

What are your suggestions for making this graph more readable?

  • is this a static or dynamic visualisation? – Benteh Mar 27 '18 at 8:18
  • 1
    @Benteh, dynamic – Daniel Mar 27 '18 at 17:07
  • 1
    If you removed 10% value you may lose lots of significant information, then your objective is not achieved by the accurate data. – NPN Mar 28 '18 at 5:05
  • @NPN, unfortunately, that is very true, and I considered the 10% purge as a compromise necessary for some decluttering. I am looking here basically for suggestions on how to avoid that (for, as you said, I fail my objective when I remove data). Do you have any suggestions? – Daniel Mar 28 '18 at 5:08
2

Use filters as much as possible to drill down the data representation to a meaningful level. Filters such as

  • States (Location)
  • Time
  • Dates
  • Years

And as you mentioned define different filters from Strongest to weak connections.

Example;

  • Strongest
  • Stronger
  • Medium
  • Slightly strong
  • weak

You can use different colors than darker or lighter in the same color. Test with users and see what type of filters can be implemented.

  • I am happy, if it works for you :). – NPN Mar 28 '18 at 21:36
0

I would just show the strongest connections, as this is the purpose of the graph as you said.

Would define a value that I would identify with "strongest connection" and hide the other connections that are inferior to that value.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.