def get_splitter(self, D):
if self.resampling_strategy_args is None:
self.resampling_strategy_args = {}
if not isinstance(self.resampling_strategy, str):
if issubclass(self.resampling_strategy, BaseCrossValidator) or\
issubclass(self.resampling_strategy, _RepeatedSplits) or\
issubclass(self.resampling_strategy, BaseShuffleSplit):
class_name = self.resampling_strategy.__name__
if class_name not in __baseCrossValidator_defaults__:
raise ValueError('Unknown CrossValidator.')
ref_arg_dict = __baseCrossValidator_defaults__[class_name]
y = D.data['Y_train'].ravel()
if class_name == 'PredefinedSplit':
if 'test_fold' not in self.resampling_strategy_args:
raise ValueError('Must provide parameter test_fold'
' for class PredefinedSplit.')
if class_name == 'LeaveOneGroupOut' or \
class_name == 'LeavePGroupsOut' or\
class_name == 'GroupKFold' or\
class_name == 'GroupShuffleSplit':
if 'groups' not in self.resampling_strategy_args:
raise ValueError('Must provide parameter groups '
'for chosen CrossValidator.')
try:
if self.resampling_strategy_args[
'groups'].shape != y.shape:
raise ValueError('Groups must be array-like '
'with shape (n_samples,).')
except Exception:
raise ValueError('Groups must be array-like '
'with shape (n_samples,).')
else:
if 'groups' in self.resampling_strategy_args:
if self.resampling_strategy_args[
'groups'].shape != y.shape:
raise ValueError('Groups must be array-like'
' with shape (n_samples,).')
# Put args in self.resampling_strategy_args
for key in ref_arg_dict:
if key == 'n_splits':
if 'folds' not in self.resampling_strategy_args:
self.resampling_strategy_args[
'folds'] = ref_arg_dict['n_splits']
else:
if key not in self.resampling_strategy_args:
self.resampling_strategy_args[key] = ref_arg_dict[
key]
# Instantiate object with args
init_dict = copy.deepcopy(self.resampling_strategy_args)
init_dict.pop('groups', None)
if 'folds' in init_dict:
init_dict['n_splits'] = init_dict.pop('folds', None)
cv = copy.deepcopy(self.resampling_strategy)(**init_dict)
if 'groups' not in self.resampling_strategy_args:
self.resampling_strategy_args['groups'] = None
return cv
y = D.data['Y_train']
shuffle = self.resampling_strategy_args.get('shuffle', True)
train_size = 0.67
if self.resampling_strategy_args:
train_size = self.resampling_strategy_args.get(
'train_size', train_size)
test_size = float("%.4f" % (1 - train_size))
if D.info['task'] in CLASSIFICATION_TASKS and \
D.info['task'] != MULTILABEL_CLASSIFICATION:
y = y.ravel()
if self.resampling_strategy in [
'holdout', 'holdout-iterative-fit'
]:
if shuffle:
try:
cv = StratifiedShuffleSplit(n_splits=1,
test_size=test_size,
random_state=1)
test_cv = copy.deepcopy(cv)
next(test_cv.split(y, y))
except ValueError as e:
if 'The least populated class in y has only' in e.args[
0]:
cv = ShuffleSplit(n_splits=1,
test_size=test_size,
random_state=1)
else:
raise e
else:
tmp_train_size = int(np.floor(train_size * y.shape[0]))
test_fold = np.zeros(y.shape[0])
test_fold[:tmp_train_size] = -1
cv = PredefinedSplit(test_fold=test_fold)
cv.n_splits = 1 # As sklearn is inconsistent here
elif self.resampling_strategy in [
'cv', 'partial-cv', 'partial-cv-iterative-fit'
]:
if shuffle:
cv = StratifiedKFold(
n_splits=self.resampling_strategy_args['folds'],
shuffle=shuffle,
random_state=1)
else:
cv = KFold(n_splits=self.resampling_strategy_args['folds'],
shuffle=shuffle,
random_state=1)
else:
raise ValueError(self.resampling_strategy)
else:
if self.resampling_strategy in [
'holdout', 'holdout-iterative-fit'
]:
# TODO shuffle not taken into account for this
if shuffle:
cv = ShuffleSplit(n_splits=1,
test_size=test_size,
random_state=1)
else:
tmp_train_size = int(np.floor(train_size * y.shape[0]))
test_fold = np.zeros(y.shape[0])
test_fold[:tmp_train_size] = -1
cv = PredefinedSplit(test_fold=test_fold)
cv.n_splits = 1 # As sklearn is inconsistent here
elif self.resampling_strategy in [
'cv', 'partial-cv', 'partial-cv-iterative-fit'
]:
cv = KFold(n_splits=self.resampling_strategy_args['folds'],
shuffle=shuffle,
random_state=1)
else:
raise ValueError(self.resampling_strategy)
return cv
def get_splitter(self, D):
if self.resampling_strategy_args is None:
self.resampling_strategy_args = {}
if not isinstance(self.resampling_strategy, str):
if issubclass(self.resampling_strategy, BaseCrossValidator) or\
issubclass(self.resampling_strategy, _RepeatedSplits) or\
issubclass(self.resampling_strategy, BaseShuffleSplit):
class_name = self.resampling_strategy.__name__
if class_name not in __baseCrossValidator_defaults__:
raise ValueError('Unknown CrossValidator.')
ref_arg_dict = __baseCrossValidator_defaults__[class_name]
y = D.data['Y_train'].ravel()
if class_name == 'PredefinedSplit':
if 'test_fold' not in self.resampling_strategy_args:
raise ValueError('Must provide parameter test_fold'
' for class PredefinedSplit.')
if class_name == 'LeaveOneGroupOut' or \
class_name == 'LeavePGroupsOut' or\
class_name == 'GroupKFold' or\
class_name == 'GroupShuffleSplit':
if 'groups' not in self.resampling_strategy_args:
raise ValueError('Must provide parameter groups '
'for chosen CrossValidator.')
try:
if self.resampling_strategy_args['groups'].shape != y.shape:
raise ValueError('Groups must be array-like '
'with shape (n_samples,).')
except Exception:
raise ValueError('Groups must be array-like '
'with shape (n_samples,).')
else:
if 'groups' in self.resampling_strategy_args:
if self.resampling_strategy_args['groups'].shape != y.shape:
raise ValueError('Groups must be array-like'
' with shape (n_samples,).')
# Put args in self.resampling_strategy_args
for key in ref_arg_dict:
if key == 'n_splits':
if 'folds' not in self.resampling_strategy_args:
self.resampling_strategy_args['folds'] = ref_arg_dict['n_splits']
else:
if key not in self.resampling_strategy_args:
self.resampling_strategy_args[key] = ref_arg_dict[key]
# Instantiate object with args
init_dict = copy.deepcopy(self.resampling_strategy_args)
init_dict.pop('groups', None)
if 'folds' in init_dict:
init_dict['n_splits'] = init_dict.pop('folds', None)
cv = copy.deepcopy(self.resampling_strategy)(**init_dict)
if 'groups' not in self.resampling_strategy_args:
self.resampling_strategy_args['groups'] = None
return cv
y = D.data['Y_train'].ravel()
shuffle = self.resampling_strategy_args.get('shuffle', True)
train_size = 0.67
if self.resampling_strategy_args:
train_size = self.resampling_strategy_args.get('train_size',
train_size)
test_size = 1 - train_size
if D.info['task'] in CLASSIFICATION_TASKS and \
D.info['task'] != MULTILABEL_CLASSIFICATION:
if self.resampling_strategy in ['holdout',
'holdout-iterative-fit']:
if shuffle:
try:
cv = StratifiedShuffleSplit(n_splits=1,
train_size=train_size,
test_size=test_size,
random_state=1)
test_cv = copy.deepcopy(cv)
next(test_cv.split(y, y))
except ValueError as e:
if 'The least populated class in y has only' in e.args[0]:
cv = ShuffleSplit(n_splits=1, train_size=train_size,
test_size=test_size, random_state=1)
else:
raise e
else:
tmp_train_size = int(np.floor(train_size * y.shape[0]))
test_fold = np.zeros(y.shape[0])
test_fold[:tmp_train_size] = -1
cv = PredefinedSplit(test_fold=test_fold)
cv.n_splits = 1 # As sklearn is inconsistent here
elif self.resampling_strategy in ['cv', 'partial-cv',
'partial-cv-iterative-fit']:
if shuffle:
cv = StratifiedKFold(
n_splits=self.resampling_strategy_args['folds'],
shuffle=shuffle, random_state=1)
else:
cv = KFold(n_splits=self.resampling_strategy_args['folds'],
shuffle=shuffle, random_state=1)
else:
raise ValueError(self.resampling_strategy)
else:
if self.resampling_strategy in ['holdout',
'holdout-iterative-fit']:
# TODO shuffle not taken into account for this
if shuffle:
cv = ShuffleSplit(n_splits=1, train_size=train_size,
test_size=test_size, random_state=1)
else:
tmp_train_size = int(np.floor(train_size * y.shape[0]))
test_fold = np.zeros(y.shape[0])
test_fold[:tmp_train_size] = -1
cv = PredefinedSplit(test_fold=test_fold)
cv.n_splits = 1 # As sklearn is inconsistent here
elif self.resampling_strategy in ['cv', 'partial-cv',
'partial-cv-iterative-fit']:
cv = KFold(n_splits=self.resampling_strategy_args['folds'],
shuffle=shuffle, random_state=1)
else:
raise ValueError(self.resampling_strategy)
return cv
