[MRG] MNT Tools for working with what's new#11800
[MRG] MNT Tools for working with what's new#11800ogrisel merged 8 commits intoscikit-learn:masterfrom
Conversation
qinhanmin2014
left a comment
There was a problem hiding this comment.
I've not gone through the script carefully but it seems good.
I somehow remember that a core dev once told contributors to reference the issue instead of the pull request but I cannot find relevant comment.
doc/whats_new/v0.20.rst
Outdated
| - |Fix| Fixed a bug in :class:`linear_model.ARDRegression` and | ||
| :class:`linear_model.BayesianRidge` which caused NaN predictions when fitted | ||
| with a constant target. | ||
| :issue:`10195` by :user:`Jörg Döpfert <jdoepfert>`. |
|
|
||
| - |Fix| Fixed a bug in :func:`manifold.spectral_embedding` where the | ||
| normalization of the spectrum was using a division instead of a | ||
| multiplication. :issue:`8129` by :user:`Jan Margeta <jmargeta>`, |
There was a problem hiding this comment.
Yeah, we're inconsistent all over the place with whether we reference PR or issue. I'm not going to be pedantic about this, especially now that I have a script (included here) which tries to break down the differences with the GitHub AI.
|
On the contrary, I tell people to reference the PR not the issue, in order to make it simple to use a script like this :) |
|
I should probably split the maint tools off this PR and just merge the what's new changes. |
|
I've merged the what's new changes and have left this open as a PR focused on tools. |
|
Will it be better to put these scripts in a separate repo (e.g., scikit-learn/maint-tools)? It seems that these scripts are for maintainers, not for users. |
|
yes, and I think it doesn't hurt to keep maintenance scripts here. but a
separate repo is okay
|
|
I have no strong opinions about having it here vs in a separate repo. |
I'll approve if you do :) So feel free to merge. |
|
Let's merge those scripts in the main repo for simplicity and discoverability. |
…)" This reverts commit e168bd7.
…)" This reverts commit e168bd7.
No description provided.