-
-
Notifications
You must be signed in to change notification settings - Fork 26k
Refactor linear model coordinate descent code to work as LARS for cv #2186
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
i'll look at this when i'm home a bit too. 2013/7/22, Alexandre Gramfort notifications@github.com:
|
@fabianp spotted a bug with precompute. |
than ``tol``. | ||
normalize : boolean, optional, default False | ||
If ``True``, the regressors X will be normalized before regression. | ||
WARNING : will be deprecated in 0.15 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Could you point me to where the decision to deprecate has been made? Can we also deprecate normalize in orthogonal matching pursuit? The code is on the point of unmaintainability because of branching in case of normalization and copy_X, copy_Gram, copy_Xy interaction.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Could you point me to where the decision to deprecate has been made?
search for :
- if normalize is not None:
-
warnings.warn("normalize param will be removed in 0.15."
-
" Intercept fitting and feature normalization will be"
-
" done in estimators.",
-
DeprecationWarning, stacklevel=2)
Can
we also deprecate normalize in orthogonal matching pursuit? The code is on
the point of unmaintainability because of branching in case of normalization
and copy_X, copy_Gram, copy_Xy interaction.
Yes I think it's the most reasonable option. Indeed all these options make
it a nightmare to maintain.
cleanup, rebased, ready for review ! warning the diff is horrible as many functions have moved. It's a big refactoring. |
travis is happy ! |
Merged by rebase |
there is a million dollar bug cf. plot_model_lasso_model_selection.py
tests pass but the example output is wrong...