On Tue, Apr 17, 2012 at 04:22:50PM +0200, Alexandre Gramfort wrote:
> what would be the semantic of scale_params?

scale_params=False by default
scale_params=True would scale the parameters by a data-dependent terms
such as C_min.

> shall we touch every estimator

No, because for some estimators we do not have a good heuristic to set
the scale. As we gain better understandings of the various estimators,
hopefully we'll come up with scale_params for them.

> or assume scale_params=True if not present as attribute?

In the GridSearchCV, we should test if an estimator has a 'scale_params'
parameter (using get_param_names) and then check if it is set to false,
and then, and only then, complain.

G

------------------------------------------------------------------------------
Better than sec? Nothing is better than sec when it comes to
monitoring Big Data applications. Try Boundary one-second 
resolution app monitoring today. Free.
http://p.sf.net/sfu/Boundary-dev2dev
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to