API add option dual="auto" for LinearSVC/LinearSVR
#24731
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
dual="auto" for LinearSVC/LinearSVR
glemaitre
reviewed
Oct 23, 2022
| @@ -37,9 +37,11 @@ class LinearSVC(LinearClassifierMixin, SparseCoefMixin, BaseEstimator): | |||
| square of the hinge loss. The combination of ``penalty='l1'`` | |||
| and ``loss='hinge'`` is not supported. | |||
|
|
|||
| dual : bool, default=True | |||
| dual : {"auto", False, True}, default= "auto" | |||
There was a problem hiding this comment.
I think that we still need to keep the default as-is and make a deprecation cycle to change the default in 1.4
Suggested change
| dual : {"auto", False, True}, default= "auto" | |
| dual : "auto" or bool, default=True |
Comment on lines
+43
to
+44
| Sets dual = False when n_samples > n_features if dual is "auto" | ||
| and dual = True otherwise. |
There was a problem hiding this comment.
Suggested change
| Sets dual = False when n_samples > n_features if dual is "auto" | |
| and dual = True otherwise. | |
| `dual="auto"` is equivalent to `dual=False` when | |
| `n_samples > n_features` and `dual=True` otherwise. |
| optimization problem. Prefer dual=False when n_samples > n_features. | ||
| optimization problem. | ||
| Sets dual = False when n_samples > n_features if dual is "auto" | ||
| and dual = True otherwise. | ||
|
|
There was a problem hiding this comment.
Suggested change
| .. versionchanged:: 1.4 | |
| The default value will change from `True` to `"auto"` in 1.4. |
| @@ -210,7 +212,7 @@ def __init__( | |||
| penalty="l2", | |||
| loss="squared_hinge", | |||
| *, | |||
| dual=True, | |||
| dual="auto", | |||
There was a problem hiding this comment.
For making the change for the future, you can check the following documentation: https://scikit-learn.org/dev/developers/contributing.html#change-the-default-value-of-a-parameter
Comment on lines
+373
to
+374
| Sets dual = False when n_samples > n_features if dual is "auto" | ||
| and dual = True otherwise. |
There was a problem hiding this comment.
Suggested change
| Sets dual = False when n_samples > n_features if dual is "auto" | |
| and dual = True otherwise. | |
| `dual="auto"` is equivalent to `dual=False` when | |
| `n_samples > n_features` and `dual=True` otherwise. | |
| .. versionchanged:: 1.4 | |
| The default value will change from `True` to `"auto"` in 1.4. |
| @@ -362,9 +367,11 @@ class LinearSVR(RegressorMixin, LinearModel): | |||
| To lessen the effect of regularization on synthetic feature weight | |||
| (and therefore on the intercept) intercept_scaling has to be increased. | |||
|
|
|||
| dual : bool, default=True | |||
| dual : {"auto", False, True}, default= "auto" | |||
There was a problem hiding this comment.
Suggested change
| dual : {"auto", False, True}, default= "auto" | |
| dual : "auto" or bool, default=True |
|
The CI tell us that we need to modify the We also need to acknowledge the change in the changelog by adding an entry in |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Reference Issues/PRs
In reference to issue #24651. The original issue is updated with the relevant benchmarks in the comments.
What does this implement/fix? Explain your changes.
Sets the default value to "auto" and selects if the problem needs to be solved in primal or dual.
Any other comments?