Ego data in GraphML output

We are beginning data analysis and have been using Gephi and R to open our Network Canvas GraphML files, but are only able to see the network data and not the answers to ‘survey-style’ questions. Is there a way we open/transform the GraphML files so we can see the ego data?

The short answer is that ego data can’t be represented in GraphML. However, GraphML files do represent ego data with our format, but the issue is that no other software can read it because it isn’t a standard feature of the format. We implemented storing ego data in a way that is compatible with GraphML standard, so Server can read the ego data stored in the files. Using Server, you can upload the existing GraphML files you have and export them as CSVs (which will include ego level data).

Just to add a bit more to what Kate put above:

  • Ego data can’t be represented in “conventional” GraphML (at least, not easily). The file format has no concept of ego level data - only graphs, nodes, and edges.
  • However, we take advantage of the fact that you can attach attributes to the graph object, and use this to store ego level data. We think this makes sense semantically, since ego attributes can be thought of as properties of the ego network (the graph) itself.
  • While that enhancement is compatible with vanilla GraphML, we made other enhancements when implementing our own file format that may not be. If you encounter a bug when loading one of our files, please let us know and we will be happy to help you work around the issue.

Further information about our GraphML based file format can be found here: Network Canvas GraphML Format - Network Canvas Documentation