tft.tfidf

Maps the terms in x to their term frequency * inverse document frequency.

tft.tfidf(
    x: tf.SparseTensor,
    vocab_size: int,
    smooth: bool = True,
    name: Optional[str] = None
) -> Tuple[tf.SparseTensor, tf.SparseTensor]

The term frequency of a term in a document is calculated as (count of term in document) / (document size)

The inverse document frequency of a term is, by default, calculated as 1 + log((corpus size + 1) / (count of documents containing term + 1)).

Example usage:

def preprocessing_fn(inputs):
  integerized = tft.compute_and_apply_vocabulary(inputs['x'])
  vocab_size = tft.get_num_buckets_for_transformed_feature(integerized)
  vocab_index, tfidf_weight = tft.tfidf(integerized, vocab_size)
  return {
     'index': vocab_index,
     'tf_idf': tfidf_weight,
     'integerized': integerized,
  }
raw_data = [dict(x=["I", "like", "pie", "pie", "pie"]),
            dict(x=["yum", "yum", "pie"])]
feature_spec = dict(x=tf.io.VarLenFeature(tf.string))
raw_data_metadata = tft.tf_metadata.dataset_metadata.DatasetMetadata(
    tft.tf_metadata.schema_utils.schema_from_feature_spec(feature_spec))
with tft_beam.Context(temp_dir=tempfile.mkdtemp()):
  transformed_dataset, transform_fn = (
      (raw_data, raw_data_metadata)
      | tft_beam.AnalyzeAndTransformDataset(preprocessing_fn))
transformed_data, transformed_metadata = transformed_dataset
transformed_data
[{'index': array([0, 2, 3]), 'integerized': array([3, 2, 0, 0, 0]),
  'tf_idf': array([0.6, 0.28109303, 0.28109303], dtype=float32)},
 {'index': array([0, 1]), 'integerized': array([1, 1, 0]),
  'tf_idf': array([0.33333334, 0.9369768 ], dtype=float32)}]

  example strings: [["I", "like", "pie", "pie", "pie"], ["yum", "yum", "pie]]
  in: SparseTensor(indices=[[0, 0], [0, 1], [0, 2], [0, 3], [0, 4],
                            [1, 0], [1, 1], [1, 2]],
                   values=[1, 2, 0, 0, 0, 3, 3, 0])
  out: SparseTensor(indices=[[0, 0], [0, 1], [0, 2], [1, 0], [1, 1]],
                    values=[1, 2, 0, 3, 0])
       SparseTensor(indices=[[0, 0], [0, 1], [0, 2], [1, 0], [1, 1]],
                    values=[(1/5)*(log(3/2)+1), (1/5)*(log(3/2)+1), (3/5),
                            (2/3)*(log(3/2)+1), (1/3)]

Note: the first doc's duplicate "pie" strings have been combined to one output, as have the second doc's duplicate "yum" strings.

Args
x	A 2D SparseTensor representing int64 values (most likely that are the result of calling compute_and_apply_vocabulary on a tokenized string).
vocab_size	An int - the count of vocab used to turn the string into int64s including any OOV buckets.
smooth	A bool indicating if the inverse document frequency should be smoothed. If True, which is the default, then the idf is calculated as 1 + log((corpus size + 1) / (document frequency of term + 1)). Otherwise, the idf is 1 +log((corpus size) / (document frequency of term)), which could result in a division by zero error.
name	(Optional) A name for this operation.

Returns
Two SparseTensors with indices [index_in_batch, index_in_bag_of_words]. The first has values vocab_index, which is taken from input x. The second has values tfidf_weight.

Raises
ValueError if x does not have 2 dimensions.