We recently introduced utils.validation._check_sample_weight which returns a validated sample_weight array.

We should use it consistently throughout the code base, instead of relying on custom and adhoc checks like check_consistent_lenght or check_array (which are now handled by _check_sample_weight).

Here's a list of the estimators/functions that could make use of it (mostly in fit or partial_fit):

• CalibratedClassifierCV
• DBSCAN
• DummyClassifier
• DummyRegressor
• BaseBagging
• BaseForest
• BaseGradientBoosting
• IsotonicRegression
• KernelRidge
• GaussianNB
• BaseDiscreteNB
• KernelDensity
• BaseDecisionTree

(I left-out the linear_model module because it seems more involved there)

Could be a decent sprint issue @amueller ?

To know where a given class is defined, use e.g. git grep -n "class DBSCAN"