HigherOrderGraph / modified MultiOrderModel - #329
vineetbansal wants to merge 21 commits into
Conversation
…legating to MultiOrderModel
M-Lampert
left a comment
There was a problem hiding this comment.
The PR looks good already. I made some comments; let me know if you have any questions. I think some of the questions that I raised should be discussed together with the others in our next meeting. I said so in the comments as well.
Co-authored-by: Moritz Lampert <[email protected]>
|
Note from 09/10 meeting - okay to restrict TODO: Vineet - add test cases for |
|
@M-Lampert - since you reviewed this last, I've incorporated the changes we discussed during our meeting.
|
|
@M-Lampert - with these changes I'm realizing that the Let's discuss this tomorrow. From Moritz - okay to remove |
A
HigherOrderGraphclass of arbitrary order, constructable from temporal_graph/event_graph/path_data. It delegates most of the work to theMultiOrderModelclass which does the iterations on the lifting (withcached=False). It is also constructable usingfrom_aggregated, and has its own.liftmethod.The
MultiOrderModelclass now hasHigherOrderGraphs in its layers.None of the tests for
MultiOrderModelneeded modifications and still pass, which is reassuring. Tests forHigherOrderModelandMultiOrderModelthat assumeHigherOrderModelin layers are coming next.This PR assumes that the
EventGraphbranch is merged, as it builds on top of it.Typical workflow using these new classes: