working on visualization challenge

So, I have over 7000 pairs of nodes in my dataset. These make up 75 tree hierarchies. There are many nodes in common between the trees.

I am trying to look at the alternatives for arranging the data so that the trees remain distinct but the common node relationships are not just a black "cloud" of lines on top of each other.

I will certainly break the trees up into smaller groupings for more focused presentations. I have this interest in showing the entire enterprise in one visual without it looking like my 4 year old grandson decided to draw on the image :lol:

Are there some tips on tuning a visualization to get maximum benefit?

Thank you

One quick idea: draw each of

One quick idea: draw each of the trees independently, cloning the nodes that are shared between trees. Pack the trees nicely into a drawing. Then add edges between all copies of a cloned node, using almost transparent edges.

The above does not take into account the topology of the shared nodes. A more sophiscated version would be to create a derived graph whose nodes are the trees, with edges between nodes if the trees share a cloned node. Create a layout for this graph with, say, neato, which should keep the trees with shared nodes close to each other. Then replace the nodes in this graph with the corresponding tree layout, and then put in the semi-transparent edges. This approach should less clutter among these added edges.

Recent comments