Closed szhorvat closed 9 months ago
Resolved in release 2.14
Is pip install hypenetx['all']
something that should be suggested?
pip install hypenetx['all']
installs Jupyter, allowing the user to open one of the tutorials in the repo; it also installs the dependencies needed for the contagion and modularity sub-modules. The documentation recommends this.
Could you please clarify where the documentation mentions this command? I can only see pip install -e .['all']
in the section about installing from source.
My impression was that the target audience of this library is not just developers, but mainly researchers. I expect most of them won't want to install from source and won't necessarily notice that pip install hypenetx['all']
is also a valid command.
I see your point about the target audience being researchers. I will update the installing from PyPi instructions to include a snippet about pip install hypernetx['all']
Have just updated installation steps on the docs: https://hypernetx.readthedocs.io/en/latest/install.html#id1
I followed the installation instructions in https://hypernetx.readthedocs.io/en/latest/install.html and installed into a new virtual environment from PyPI using
pip install hypernetx
. When I load the package, the following message is shown:This is not an ideal user experience, and is likely confusing to some new users. They may be wondering: "Did I install correctly? Will some essential functionality not be working?"
I suggest:
hypernetx.algorithms.hypergraph_modularity
instead of on package load? This is a more common practice.Ref: https://github.com/openjournals/joss-reviews/issues/6016