setup test error (networkx): TypeError: add_node() takes 2 positional arguments but 3 were given

[victoriastuart]

...
nx_graph = nx.Graph()

    for node in G.node_iterator():

      nx_graph.add_node(node, G.get_node_attributes(node))

E TypeError: add_node() takes 2 positional arguments but 3 were given

halp/utilities/undirected_graph_transformations.py:71: TypeError

...

[john-arul1]

read as "takes 1 positional argument but 2 given", (discount self argument in the class def). Attributes have to be given as key word arguments
nx_graph.add_node(node)
or
nx_graph.add_node(node,weight=0.5)

[mohawk2]

This is because networkx 2.0 changed its API. PR #30 fixes this.

@john-arul1 To be specific about the error Victoria was seeing, what was needed was to change the dict argument to **kwargs arguments.

@victoriastuart You'll be able to clone my fork of this repo and make hypergraph stuff. In a perfect world, the PR will be merged and a new version released to PyPi.

@annaritz @tmmurali @bavent Any chance of the above happening? :-)

[tmmurali]

@mohawk2 Thank you very much for this PR.

@annaritz or @bavent will you be able to check the PR and merge? I took a look at the changes and they seem good to me.

[mohawk2]

@annaritz Thank you for merging the PR! https://pypi.org/project/halp/ looks like a new release hasn't been cut yet (current one is from @bavent from 2014). I'll be pleased to assist if that would be useful.

[tmmurali]

@mohawk2 I have created a new release on GitHub. I have asked @bavent to transfer ownership of the halp page on pypi.org to me, if that is at all possible.

[mohawk2]

@tmmurali Thank you for rapid response! https://pypi.org/help/#request-ownership suggests it absolutely is possible.

[tmmurali]

@mohawk2 thanks for this pointer. I should have checked myself! I will wait to hear from @bavent first.

[mohawk2]

If I had (insert currency unit) for each time I asked something that I could theoretically have found out myself... I'm not judging! :-D