Re-wrote "How to debug a model"

[kuvychko]

References

• Towards how to debug a model #109

• References Project Board: https://github.com/pymc-devs/pymc-examples/projects/1

• Notebook follows style guide https://docs.pymc.io/en/latest/contributing/jupyter_style.html

• PR description contains a link to the relevant issue: a tracker one for existing notebooks or a proposal one for new notebooks

• Check the notebook is not excluded from any pre-commit check: https://github.com/pymc-devs/pymc-examples/blob/main/.pre-commit-config.yaml

Helpful links

• https://github.com/pymc-devs/pymc-examples/blob/main/CONTRIBUTING.md

Notes for the Reviewer

• I asked on Discourse for better examples, but so far I got no replies. My examples are definitely not ideal, having a realistic model failing mid-run would work much better but I haven't figured out how to achieve that. https://discourse.pymc.io/t/example-of-using-aesara-printing-print-for-model-troubleshooting/10042

#DataUmbrellaPyMCSprint

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[twiecki]

These are supposed to reflect when it was originally written and by whom. If you find the date in the git history it should go here, same with the author (although in this case I'm pretty sure it was me). Usually upgrading/porting a NB does not add to authors but you added new information here so you should be co-author.

Reply via ReviewNB

[kuvychko]

@twiecki - thanks! I asked who authored the original in the issue tracker but didn't get a reply, I didn't think about commit history. the last commit I found was by MarcoGorelli committed on Dec 14, 2020, but it was a move from pymc3. I can't figure out how to track the original. I can put your name and Dec 14, 2020 as the date, or let me know if I should use an earlier date. I don't think I added much, just re-dressed the ideas that you put in (but figured that folks should know whom to blame since I couldn't find the original author :-) ).

[kuvychko]

@twiecki - new changes are in, thanks to @reshamas who found the commit history.

[reshamas]

@kuvychko Thanks for this PR. Some comments below:

a) For future, I think it would be better to submit notebooks in a separate PR, no?
Unless for some reason these notebooks are connected. No need to change it now.

b) The description of this PR needs some text.
Here is an example of what can be added:
#401

c) The debug notebook has this issue opened by @OriolAbril:
#109

Towards #109

d) For the history of the debugging notebook, I did this search:
https://github.com/pymc-devs/pymc/commits/76bd40e0d1db4fd906b7d558dcf1bd79f1011963/docs/source/notebooks/howto_debugging.ipynb

e) For the history of the spline notebook, this looks like the start of it:
#230

#DataUmbrellaPyMCSprint
@symeneses

[kuvychko]

@reshamas - thank you, really appreciate your comments and help! I made the changes accordingly.

[OriolAbril]

went over style only for now

[kuvychko]

@OriolAbril - thanks for the detailed style review and comments, super helpful! I committed requested changes. Thanks!

[twiecki]

@OriolAbril Do you know what's up with rtd?

[OriolAbril]

I haven't checked, but I'd guess it is related to #409. That is, the PR was opened before the error happened and needs a rebase to include the version pins, otherwise the preview installes the latest and non-working versions.

[kuvychko]

@OriolAbril - any actions needed on my part to get this PR merged? Also, anything I need to do differently for future PRs to avoid this?

[symeneses]

@OriolAbril - any actions needed on my part to get this PR merged? Also, anything I need to do differently for future PRs to avoid this?

You need to rebase upstream to origin, so that your fork is updated.

# update you main branch
git checkout main
git fetch upstream
git rebase upstream/main
# update your branch
git checkout feature
git rebase main

If this works 🤞🏽 , you should see these 2 commits when executing git log:

commit 76fae3ffff138021abf484ea4317e74d31f970cb (HEAD -> feature, upstream/main, origin/main, origin/HEAD, main)
Author: Reshama Shaikh <reshama.stat@gmail.com>
Date:   Thu Aug 4 12:17:04 2022 -0400

commit 50db8524366302c262f08505c670b995b5b88438
Author: Sandra Yojana Meneses <sandrayojana@gmail.com>
Date:   Wed Aug 3 09:57:04 2022 +0200

    Upgrade notebook Gaussian Process to V4 (#404)

Once you do this, push again the changes in your branch.

After you fork a repo and before your PR is closed, new PRs can be merged to main, normally, that shouldn't affect your work but in this case, the building of docs was failing due to some issues in the new versions which was fixed in a different PR as Oriol explained. So now, this PR needs to have the last update so that the docs can be built. A way to avoid this is to update your fork right before creating the PR so that your branch has the latest updates.

[kuvychko]

@symeneses - thank you!!! looks like it worked 👍

[kuvychko]

@OriolAbril - any issues with the merge? I wanna make sure it doesn't fall through the cracks somehow. Thanks!