On Thu, Mar 22, 2012 at 11:47:02AM +0000, Conrad Lee wrote:
> Yes, that makes sense, and I tried that. But any RFECV expect every
> estimator to have a .coef_ attribute (it recursively eliminates features
> based on their coefficients).
Good point,
I see 2 options:
1. Hack: Subclass Gridsearch so that it sets
self.coef_ = self.best_estimator.coef_ at the end of the fit,
2. Better solution: send us a pull request to try and use
estimator.best_estimator.coefs_ if estimator has not attribute
coefs_ in the RFE and RFECV object.
Cheers,
Gaël
------------------------------------------------------------------------------
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________
Scikit-learn-general mailing list
Scikit-learn-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
