def _validate_linear_feature_columns(self, features):
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(features)
self._feature_columns_inferred = True
elif self._feature_columns_inferred:
this_dict = {c.name: c for c in self._linear_feature_columns}
that_dict = {
c.name: c for c in layers.infer_real_valued_columns(features)
}
if this_dict != that_dict:
raise ValueError(
"Feature columns, expected %s, got %s.", (this_dict, that_dict))
def _validate_linear_feature_columns(self, features):
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(features)
self._feature_columns_inferred = True
elif self._feature_columns_inferred:
this_dict = {c.name: c for c in self._linear_feature_columns}
that_dict = {
c.name: c for c in layers.infer_real_valued_columns(features)
}
if this_dict != that_dict:
raise ValueError(
"Feature columns, expected %s, got %s.", (this_dict, that_dict))
def _get_train_ops(self, features, targets):
"""See base class."""
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(
features)
if not isinstance(self._linear_optimizer,
sdca_optimizer.SDCAOptimizer):
return super(LinearClassifier,
self)._get_train_ops(features, targets)
# SDCA currently supports binary classification only.
if self._n_classes > 2:
raise ValueError(
"SDCA does not currently support multi-class classification.")
global_step = contrib_variables.get_global_step()
assert global_step
logits, columns_to_variables, _ = layers.weighted_sum_from_feature_columns(
columns_to_tensors=features,
feature_columns=self._linear_feature_columns,
num_outputs=self._num_label_columns(),
weight_collections=[self._linear_weight_collection],
name="linear")
with ops.control_dependencies([self._centered_bias()]):
loss = self._loss(logits, targets,
self._get_weight_tensor(features))
logging_ops.scalar_summary("loss", loss)
train_ops = self._linear_optimizer.get_train_step(
self._linear_feature_columns, self._weight_column_name,
"logistic_loss", features, targets, columns_to_variables,
global_step)
return train_ops, loss
def _get_train_ops(self, features, targets):
"""See base class."""
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(features)
if not isinstance(self._linear_optimizer, sdca_optimizer.SDCAOptimizer):
return super(LinearClassifier, self)._get_train_ops(features, targets)
# SDCA currently supports binary classification only.
if self._n_classes > 2:
raise ValueError(
"SDCA does not currently support multi-class classification.")
global_step = contrib_variables.get_global_step()
assert global_step
logits, columns_to_variables, _ = layers.weighted_sum_from_feature_columns(
columns_to_tensors=features,
feature_columns=self._linear_feature_columns,
num_outputs=self._num_label_columns(),
weight_collections=[self._linear_weight_collection],
name="linear")
with ops.control_dependencies([self._centered_bias()]):
loss = self._loss(logits, targets, self._get_weight_tensor(features))
logging_ops.scalar_summary("loss", loss)
train_ops = self._linear_optimizer.get_train_step(
self._linear_feature_columns, self._weight_column_name, "logistic_loss",
features, targets, columns_to_variables, global_step)
return train_ops, loss
def _get_train_ops(self, features, targets):
"""See base class."""
if isinstance(self._linear_optimizer, sdca_optimizer.SDCAOptimizer):
raise ValueError("SDCAOptimizer does not currently support regression.")
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(features)
return super(LinearRegressor, self)._get_train_ops(features, targets)
def _get_train_ops(self, features, targets):
"""See base class."""
if isinstance(self._linear_optimizer, sdca_optimizer.SDCAOptimizer):
raise ValueError(
"SDCAOptimizer does not currently support regression.")
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(
features)
return super(LinearRegressor, self)._get_train_ops(features, targets)
def infer_real_valued_columns_from_input_fn(input_fn):
"""Creates `FeatureColumn` objects for inputs defined by `input_fn`.
This interprets all inputs as dense, fixed-length float values. This creates
a local graph in which it calls `input_fn` to build the tensors, then discards
it.
Args:
input_fn: Function returning a tuple of input and target `Tensor` objects.
Returns:
List of `FeatureColumn` objects.
"""
with ops.Graph().as_default():
features, _ = input_fn()
return layers.infer_real_valued_columns(features)
def infer_real_valued_columns_from_input_fn(input_fn):
"""Creates `FeatureColumn` objects for inputs defined by `input_fn`.
This interprets all inputs as dense, fixed-length float values. This creates
a local graph in which it calls `input_fn` to build the tensors, then discards
it.
Args:
input_fn: Function returning a tuple of input and target `Tensor` objects.
Returns:
List of `FeatureColumn` objects.
"""
with ops.Graph().as_default():
features, _ = input_fn()
return layers.infer_real_valued_columns(features)
def _get_train_ops(self, features, targets):
"""See base class."""
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(features)
return super(LinearClassifier, self)._get_train_ops(features, targets)
def _get_train_ops(self, features, targets):
"""See base class."""
if self._dnn_feature_columns is None:
self._dnn_feature_columns = layers.infer_real_valued_columns(features)
return super(DNNRegressor, self)._get_train_ops(features, targets)
def _get_train_ops(self, features, targets):
"""See base class."""
if self._linear_feature_columns is None:
self._linear_feature_columns = layers.infer_real_valued_columns(
features)
return super(LinearClassifier, self)._get_train_ops(features, targets)
