Different ways of handling large Graphs?

Question

A single GraphML files contains over ~100,000 Nodes/entities (with parent child relationship). I don’t want to display all the Nodes at once. For this project, I would only like to display a subset of the graph, and based on the user event display the successive nodes for the corresponding Node the user clicked on.

With reference to your website, we found, Balloon layout can handle huge graphs, but I believe, it load ever thing at once.  What are my other possible options?

if possible, could you provide us with a sample, or any reference material..

Answer 1

You are right, yEd always loads the whole graph at once. And honestly, it is simply not designed to handle graphs with that many nodes.

If you have to work with large sets of data, please consider developing a custom application using one of the yFiles diagramming libraries which are available for many platforms - see http://www.yworks.com/ for information on the yFiles product family. E.g. yFiles for Java comes with small sample applications that demonstrate possible approaches to filtering large data sets (demo.layout.multipage.MultiPageLayoutDemo, demo.view.graphexplorer.GraphExplorerDemo, and to some extent  emo.view.advanced.ModelViewManagerDemo).