Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Installation issues #11

Closed
pmli opened this issue Jan 31, 2022 · 2 comments
Closed

Installation issues #11

pmli opened this issue Jan 31, 2022 · 2 comments

Comments

@pmli
Copy link

pmli commented Jan 31, 2022

I found a few issues with the installation instructions in README.md:

  1. It seems that the minimal required Python version is 3.5 (the @ operator is used in linear_model/_linear_model.py).
  2. Running pip install pynumdiff in a fresh virtual environment brings up issues with pychebfun for me. It seems that numpy, scipy, and matplotlib need to be installed prior to installing pychebfun (and pynumdiff).
  3. Regarding installing from source, I heard somewhere that pip install . is recommended over python setup.py install. And, indeed, running the latter raises SetuptoolsDeprecationWarning and EasyInstallDeprecationWarning for me.

(This is related to my JOSS review: openjournals/joss-reviews#4078)

@florisvb
Copy link
Owner

florisvb commented Feb 1, 2022

Thanks for identifying these issues. I have attempted to resolve all of them through various methods.

  1. I removed the @ operator (replaced with cvxpy.matmul), though I also added python>=3.5 as a requirement, though in principle most of the code should work on python 2.7.
  2. I removed pychebfun and cvxpy as requirements, and added them as optional requirements. This eliminates that particular issue, furthermore, pychebfun is only required for one option, and there is no need to require all users to have this installed.
  3. It is challenging to meet all the installation requirements (especially on an older system like Ubuntu 16.04). To help with this I added more detailed install instructions include several examples that work through a complete install in an empty virtual environment.

Let me know if these changes resolve the issues raised.

@pmli
Copy link
Author

pmli commented Feb 1, 2022

Ok, thanks, I guess this resolves the issues, so I'll close this. I would just add some comments:

  1. From using vermin, it seems that the minimum required Python version is 3.3. (I do not mean to suggest with this that PyNumDiff should support Python versions that had reached their end of life.)
  2. Sounds good.
  3. When I had been using Ubuntu 16.04, I had a good experience using pyenv to install newer versions of Python without admin privileges. Does this help you to avoid the installation instructions for these cases?

@pmli pmli closed this as completed Feb 1, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants