Construct a Categorical Column with In-Memory Vocabulary

Use this when your inputs are in string or integer format, and you have an in-memory vocabulary mapping each value to an integer ID. By default, out-of-vocabulary values are ignored. Use default_value to specify how to include out-of-vocabulary values. For the input dictionary features, features$key is either tensor or sparse tensor object. If it's tensor object, missing values can be represented by -1 for int and '' for string.

column_categorical_with_vocabulary_list(..., vocabulary_list,
  dtype = NULL, default_value = -1L, num_oov_buckets = 0L)

Arguments

...

Expression(s) identifying input feature(s). Used as the column name and the dictionary key for feature parsing configs, feature tensors, and feature columns.

vocabulary_list

An ordered iterable defining the vocabulary. Each feature is mapped to the index of its value (if present) in vocabulary_list. Must be castable to dtype.

dtype

The type of features. Only string and integer types are supported. If NULL, it will be inferred from vocabulary_list.

default_value

The value to use for values not in vocabulary_list.

num_oov_buckets

Non-negative integer, the number of out-of-vocabulary buckets. All out-of-vocabulary inputs will be assigned IDs in the range [vocabulary_size, vocabulary_size+num_oov_buckets) based on a hash of the input value. A positive num_oov_buckets can not be specified with default_value.

Value

A categorical column with in-memory vocabulary.

Details

Note that these values are independent of the default_value argument.

Raises

  • ValueError: if vocabulary_list is empty, or contains duplicate keys.

  • ValueError: if dtype is not integer or string.

See also