Module: tf.feature_column

Defined in tensorflow/python/feature_column/feature_column_lib.py.

FeatureColumns: tools for ingesting and representing features.

Functions

bucketized_column(...): Represents discretized dense input.

categorical_column_with_hash_bucket(...): Represents sparse feature where ids are set by hashing.

categorical_column_with_identity(...): A _CategoricalColumn that returns identity values.

categorical_column_with_vocabulary_file(...): A _CategoricalColumn with a vocabulary file.

categorical_column_with_vocabulary_list(...): A _CategoricalColumn with in-memory vocabulary.

crossed_column(...): Returns a column for performing crosses of categorical features.

embedding_column(...): _DenseColumn that converts from sparse, categorical input.

indicator_column(...): Represents multi-hot representation of given categorical column.

input_layer(...): Returns a dense Tensor as input layer based on given feature_columns.

linear_model(...): Returns a linear prediction Tensor based on given feature_columns.

make_parse_example_spec(...): Creates parsing spec dictionary from input feature_columns.

numeric_column(...): Represents real valued or numerical features.

shared_embedding_columns(...): List of dense columns that convert from sparse, categorical input.

weighted_categorical_column(...): Applies weight values to a _CategoricalColumn.