Closed fbergmann closed 2 years ago
Sounds good to me! Would you suggest to group by application or specific software? For instance, when PyEnzyme is used for modelling the command pip install pyenzyme[modelling]
would install Copasi and PySCeS.
The 'requirements.txt' file is specific for the REST-API, but I'd change the Dockerfile to not rely on that.
The current requirements in the txt file and setup.py are out of date. (PyEnzyme wont run without plotly and seaborn). It would be good to update those. I recommend removing the requirements file and just doing it in the setup.py file.
Ideally there would be different extras to keep the core library dependencies as small as possible. WIth the library falling back to None object if a certain dependency is not available. Possible extras are:
having then an 'all' that combine the individual extras. That way people just needing core functionality could install
pip install pyenzyme
others wouldpip install pyenzyme[all]
.