https://topsbm.github.io/ implementation not compatible with gt 2.41
Python version: 3.9 graph-tool 2.41, from conda-forge, downloaded today July,1 2021
Hi, not sure if this is the right place for the bug, since it concerns a graph-tool application from https://topsbm.github.io/ but given the coauthorship on the reference paper, and the apparent graph-tool dependency issue, I thought of asking here too (also opened an issue on https://github.com/martingerlach/hSBM_Topicmodel/issues/35) Apologies in advance if the request is not pertinent, and feel free to close the issue if this is the case. Any help at all with this issue would be greatly appreciated!
I was running the example notebook to do topic modelling on a collection of documents, including the hyperlinks layer.
When fitting the model, the function gt.minimize_nested_blockmodel_dl(self.g)#, deg_corr=True, layers = True, state_args=state_args)
from graph-tool is called with these arguments.
The error message simply says that deg_corr
and layers
are not arguments of the function
If I looked at the right snippets in the source code, it seems these arguments have been removed in version 2.38
Maybe a simple solution could be to revert to an older version of graph-tool? Could this cause other problems?
It's my first approach to SBM at all, so I was wondering whether this is all expected, and how to fix this error in order to be able to apply the methodology you described in https://epjdatascience.springeropen.com/articles/10.1140/epjds/s13688-021-00288-5 . Also, congrats for the article. It really looks awesome, and I look forward to apply the methodology to our use case.