Closed bvarjavand closed 4 years ago
unfortunately due to the attributes issue (#66), all of the datasets that are both a) edgelist and b) have node attributes will have a terrible perofrmance, because attributes cannot be incorporated into the embedding properly. this includes all CD and most of VC.
thus, we have the following options as our pipelines to pitch to a TA2 team:
we now also have seeded graph matching, though it's a little skech because it is a one-dataset pipeline and i would want to keep that fact under brians radar until we break it down into separate pieces and possibly also merge it with a graph reader.
ta2 team has made an lse-gclass pipeline from our primtives.
closing.
talk to daragh & co