tff.analytics.heavy_hitters.iblt.IbltTensorEncoder

Encodes the strings into an IBLT data structure.

Inherits From: IbltEncoder

tff.analytics.heavy_hitters.iblt.IbltTensorEncoder(
    value_shape: Sequence[int], *args, **kwargs
)

value_shape Shape of the values.
*args See IbltEncoder.
**kwargs See IbltEncoder.

Methods

compute_chunks

View source

compute_chunks(
    input_strings
)

Returns Tensor containing integer chunks for input strings.

Args
input_strings A tensor of strings.

Returns
A 2D tensor with rows consisting of integer chunks corresponding to the string indexed by the row and a trimmed input_strings that can fit in the IBLT.

compute_iblt

View source

@tf.function
compute_iblt(
    input_strings: tf.Tensor, input_values: tf.Tensor
) -> tuple[tf.Tensor, tf.Tensor]

Returns Tensor containing the values of the IBLT data structure.

Args
input_strings A 1D tensor of strings.
input_values A tensor of shape (num_input_strings, value_shape) containing values for each string.

Returns
A tuple of tensors where the first one is of shape [repetitions, table_size, num_chunks+2] whose value at index (r, h, c) corresponds to chunk c of the keys if c < num_chunks, to the counts if c == num_chunks, and to the checks if c == num_chunks + 1. The second one is of shape [repetitions, table_size, product(value_shape)] and contains the tensor values at each key.