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

Remove setuptools workaround #40

Merged
merged 1 commit into from
Aug 6, 2022
Merged

Conversation

maresb
Copy link
Contributor

@maresb maresb commented Aug 5, 2022

@maresb
Copy link
Contributor Author

maresb commented Aug 5, 2022

I seem to have recent acquired the habit of reverting your changes. 😆

@michaelosthege
Copy link
Member

I seem to have recent acquired the habit of reverting your changes. 😆

That's okay 😅 Thanks for opening the PR - I already forgot about this...

By the way, today I realized that PyMC is still on an "outdated" Aesara version and the import fixes have not yet arrived for our main branch..

@maresb
Copy link
Contributor Author

maresb commented Aug 6, 2022

Hmm, am I missing something, or are you looking at an old commit?

Aesara is at 2.7.9:
image

PyMC is on Aesara 2.7.9:
image

The upstream fix aesara-devs/aesara#1050 was merged in Aesara v2.7.8

So it seems to me like PyMC is fine, no?

@maresb
Copy link
Contributor Author

maresb commented Aug 6, 2022

Ah... do you mean that mcbackend is on an old pymc version which, when installed with PyPI would not incorporate conda-forge/conda-forge-repodata-patches-feedstock#307, and thus still be vulnerable to the issue?

Indeed one should probably wait until both PyMC and Aesara have new releases before merging this.

@michaelosthege
Copy link
Member

You're right, I'm mixing up two different things here.
I will merge this PR because the setuptools things is indeed fixed.

I was concerned about the import mess, which I just confirmed is still happening with Aesara 2.7.9 👇

image

@michaelosthege michaelosthege merged commit 3973b28 into pymc-devs:main Aug 6, 2022
@maresb
Copy link
Contributor Author

maresb commented Aug 6, 2022

Oh no!!! 😞

Which commands did you use to install?

@maresb maresb deleted the patch-1 branch August 6, 2022 09:48
@maresb
Copy link
Contributor Author

maresb commented Aug 6, 2022

Thanks for the merge at PyMC.

Does the following logic seem correct?

In the screenshot above, you're running pymc 4.1.4, but since it didn't exist on conda-forge (and won't exist until the CDN updates), it couldn't have been installed from the pymc CF package, so it's not surprising that g++ is broken.

@michaelosthege
Copy link
Member

Nevermind, I created several envs with Python 3.9, 3.10 and PyMC 4.1.2, 4.1.3 and 4.1.4 and none of the new environments had the issue.

mamba create -n pmtest412 -c conda-forge "python==3.9" "pymc==4.1.2"

After running mamba install -c conda-forge "pymc==4.1.4" in the environment that produced the screenshot above, also there the problems magicly went away.

@maresb
Copy link
Contributor Author

maresb commented Aug 6, 2022

😅 Great!!!

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

Successfully merging this pull request may close these issues.

2 participants