def fit_weights(self,
X,
y,
sample_weight=None,
parallel_profile=None,
features=None):
if features is not None:
for name, estimator in self.items():
if estimator.features is not None:
print('Overwriting features of estimator ' + name)
self[name].set_params(features=features)
# allow specifying different weights for each classifier
if isinstance(sample_weight, OrderedDict):
sample_weight = list(sample_weight.values())
else:
sample_weight = [sample_weight] * len(self)
start_time = time.time()
result = utils.map_on_cluster(parallel_profile, train_estimator,
list(self.keys()), list(self.values()),
[X] * len(self), [y] * len(self),
sample_weight)
for status, data in result:
if status == 'success':
name, estimator, spent_time = data
self[name] = estimator
print('model {:12} was trained in {:.2f} seconds'.format(
name, spent_time))
else:
print('Problem while training on the node, report:\n', data)
print("Totally spent {:.2f} seconds on training".format(time.time() -
start_time))
return self
def fit(self,
X,
y,
sample_weight=None,
parallel_profile=None,
features=None):
"""
Train all estimators on the same data.
:param X: pandas.DataFrame of shape [n_samples, n_features] with features
:param y: array-like of shape [n_samples] with labels of samples
:param sample_weight: weights of events,
array-like of shape [n_samples] or None if all weights are equal
:param features: features to train estimators
If None, estimators will be trained on `estimator.features`
:type features: None or list[str]
:param parallel_profile: profile of parallel execution system or None
:type parallel_profile: None or str
:return: self
"""
if features is not None:
for name, estimator in self.items():
if estimator.features is not None:
print('Overwriting features of estimator ' + name)
self[name].set_params(features=features)
start_time = time.time()
labels = []
for key in self.keys():
labels.append((y == names_labels_correspondence[key]) * 1)
result = map_on_cluster(parallel_profile, train_estimator,
list(self.keys()), list(self.values()),
[X] * len(self), labels,
[sample_weight] * len(self))
for status, data in result:
if status == 'success':
name, estimator, spent_time = data
self[name] = estimator
print('model {:12} was trained in {:.2f} seconds'.format(
name, spent_time))
else:
print('Problem while training on the node, report:\n', data)
print("Totally spent {:.2f} seconds on training".format(time.time() -
start_time))
return self
def fit(self, X, y, sample_weight=None):
"""
Train the classifier, will train several base classifiers on overlapping
subsets of training dataset.
:param X: pandas.DataFrame of shape [n_samples, n_features]
:param y: labels of events - array-like of shape [n_samples]
:param sample_weight: weight of events,
array-like of shape [n_samples] or None if all weights are equal
"""
if hasattr(self.base_estimator, 'features'):
assert self.base_estimator.features is None, \
'Base estimator must have None features! Use features parameter in Folding instead'
self.train_length = len(X)
group_column, (X, y, sample_weight) = self._prepare_data(X, y, sample_weight)
folds_column = self._get_folds_column(len(X), group_column)
for _ in range(self.n_folds):
self.estimators.append(clone(self.base_estimator))
if sample_weight is None:
weights_iterator = [None] * self.n_folds
else:
weights_iterator = (sample_weight[folds_column != index] for index in range(self.n_folds))
result = map_on_cluster(self.parallel_profile, train_estimator,
range(len(self.estimators)),
self.estimators,
(X.iloc[folds_column != index, :].copy() for index in range(self.n_folds)),
(y[folds_column != index] for index in range(self.n_folds)),
weights_iterator)
for status, data in result:
if status == 'success':
name, classifier, spent_time = data
self.estimators[name] = classifier
else:
print('Problem while training on the node, report:\n', data)
return self
def fit(self, X, y, sample_weight=None, parallel_profile=None, features=None):
"""
Train all estimators on the same data.
:param X: pandas.DataFrame of shape [n_samples, n_features] with features
:param y: array-like of shape [n_samples] with labels of samples
:param sample_weight: weights of events,
array-like of shape [n_samples] or None if all weights are equal
:param features: features to train estimators
If None, estimators will be trained on `estimator.features`
:type features: None or list[str]
:param parallel_profile: profile of parallel execution system or None
:type parallel_profile: None or str
:return: self
"""
if features is not None:
for name, estimator in self.items():
if estimator.features is not None:
print('Overwriting features of estimator ' + name)
self[name].set_params(features=features)
start_time = time.time()
labels = []
for key in self.keys():
labels.append((y == names_labels_correspondence[key]) * 1)
result = map_on_cluster(parallel_profile, train_estimator, list(self.keys()), list(self.values()),
[X] * len(self), labels, [sample_weight] * len(self))
for status, data in result:
if status == 'success':
name, estimator, spent_time = data
self[name] = estimator
print('model {:12} was trained in {:.2f} seconds'.format(name, spent_time))
else:
print('Problem while training on the node, report:\n', data)
print("Totally spent {:.2f} seconds on training".format(time.time() - start_time))
return self
def fit(self, X, y, parallel_profile=None, **params):
if self.multi_mode:
self.base_estimator.fit(X, y, **params)
else:
start_time = time.time()
labels = []
if len(self.models) == 0:
keys = numpy.unique(y)
for key in keys:
labels.append((y == key) * 1)
self.models[key] = clone(self.base_estimator)
else:
for key in self.models.keys():
labels.append((y == key) * 1)
sample_weight = numpy.ones(
len(X)
) if 'sample_weight' not in params else params['sample_weight']
result = map_on_cluster(parallel_profile, train_estimator,
list(self.models.keys()),
list(self.models.values()),
[X] * len(self.models), labels,
[sample_weight] * len(self.models))
for status, data in result:
if status == 'success':
name, estimator, spent_time = data
self.models[name] = estimator
print('model {:12} was trained in {:.2f} seconds'.format(
name, spent_time))
else:
print('Problem while training on the node, report:\n',
data)
print(
"Totally spent {:.2f} seconds on training".format(time.time() -
start_time))
return self
