ENH NonBLASDotWarning -> EfficiencyWarning; Improve error message
DOC Add exceptions module to modules/classes.rst
MAINT Move ConvergenceWarning, UndefinedMetricWarning et al into exceptions
MAINT Remove ChangedBehaviorWarning from base
DOC/FIX Improve DataConversionWarning's docstring
MAINT Remove sequence of sequence support from datasets
MAINT Remove return_indicator param
MAINT Remove multilabel-seq test in OVR
MAINT Remove multilable-seq test in check_cv
MAINT Remove multilabel seq test in label_binarizer
TST type_of_target returns "unknown" for multilabel-sequence types
TST _check_targets should raise a ValueError
DOC show multilabel indicator as an example; remove return_indicator param
DOC use consistent lower case y for target
and fixes a few other doc problems while I was at it.
The main take-home message is to remember that you need a backslash when
your parameter type description exceeds one line.
The Brier score allows to evaluate the quality of the predicted probabilities of a classifier. It is defined as the mean squared difference between (1) the predicted probability assigned to the possible outcomes for an item and (2) the actual outcome.
The first example did not demonstrate average=None, it now does.
The second example demonstrates that the values do not have to be
integers, they can be strings, and it also demonstrates that the order
of the results is determined by the labels argument.
the names are now consistent across methods
In LinearSVC:
'l1' -> 'hinge'
'l2' -> 'squared_hinge'
In LinearSVR:
'l1' -> 'epsilon_insensitive'
'l2' -> 'squared_epsilon_insensitive'
np.bincount raises an Exception with empty input arrays for numpy versions < 1.6.2
Add utils.fixes.bincount to tackle this issue and use it instead of np.bincount
Alexandre Gramfort