DOC use notebook-style for plot_bayesian_ridge.py #22794
Conversation
fkaren27
changed the title
lesteve
changed the title
|
Thanks for your PR, FYI I edited the title of your PR so that it contains the example filename. The reason for this is that we have a script that updates the #22406 automatically and relies on the example filename being in the PR title. |
ogrisel
left a comment
ogrisel
left a comment
There was a problem hiding this comment.
Thanks for the PR @fkaren27!
I had a look at the rendering from the CI (see the "doc artifact" link):
https://181160-843222-gh.circle-artifacts.com/0/doc/auto_examples/linear_model/plot_bayesian_ridge.html
I think the 3 matplotlib figures would benefit from being each in their own cell. So please introduce new cell markers # %% to separate them.
Note that when doing so, the final line of the first cell will be:
plt.legend(loc="best", prop=dict(size=12)the returned value of that function call will then be displayed in a yellow rectangle used to render the output of the cell in notebook-style examples. Since the string representation of a matplotlib legend object is not interesting to the reader, we silence such noisy output with a dummy variable assignment:
_ = plt.legend(loc="best", prop=dict(size=12)
You can also drop the final plt.show(), I think.
|
The black formatter complains on the CI. To automatically fixes those, you can run: or directly using VS Code ( |
jeremiedbb
left a comment
jeremiedbb
left a comment
There was a problem hiding this comment.
Thanks @fkaren27. I think the rendering would be improved by making the comments separating sections as titles
Co-authored-by: Jérémie du Boisberranger <[email protected]>
|
Merging, thanks a lot! |
Co-authored-by: Jérémie du Boisberranger <[email protected]> Co-authored-by: Loïc Estève <[email protected]>
4 participants
Reference Issues/PRs
towards #22406
This is a work in progress as part of the #pariswimlds sprint.