ENH follow-up style improvements of DET curves #10591#18169
ENH follow-up style improvements of DET curves #10591#18169lorentzenchr merged 9 commits intoscikit-learn:masterfrom
Conversation
|
ping @cmarmo @agramfort @jnothman who reviewed #10591 |
|
We should open an issue to add the |
NicolasHug
left a comment
There was a problem hiding this comment.
nits but lgtm, thanks @glemaitre
| @@ -1,7 +1,7 @@ | |||
| """ | |||
| ======================================= | |||
| ==================================== | |||
There was a problem hiding this comment.
should this be in metric instead of model_selection ?
There was a problem hiding this comment.
The plot_roc is also in this section. I am not sure what is best. I think that the idea is to compare model and select the best one in some sense. So it is not so wrong to put it in this section.
|
There's also the "False Negative Rate" ylabel that overlaps with the left plot in the example. |
|
ping @lorentzenchr @rth as well :P |
lorentzenchr
left a comment
There was a problem hiding this comment.
Don't know how to add a review comment to the following:
- In the docstring of
detection_error_tradeoff_curve, replace "Note: This metrics is used for ranking evaluation of a binary" by "This metric is used for evaluation of ranking and error tradeoffs of a binary". - In the User Guide, can we replace the word "ordinate" by "y-axis"?
|
@lorentzenchr Done |
Co-authored-by: Christian Lorentzen <lorentzen.ch@gmail.com>
|
Good to be merged :) |
…ikit-learn#18169) * follow-up on scikit-learn#10591 * DOC improvements * TST add additional test for pos_label Co-authored-by: Christian Lorentzen <lorentzen.ch@gmail.com>
Follow-up of #10591.
Closes #18174.