TF 2.0 is out! Get hands-on practice at TF World, Oct 28-31. Use code TF20 for 20% off select passes. Register now

tfp.experimental.substrates.numpy.math.linalg.sparse_or_dense_matvecmul

View source on GitHub

Returns (batched) matmul of a (sparse) matrix with a column vector.

tfp.experimental.substrates.numpy.math.linalg.sparse_or_dense_matvecmul(
    sparse_or_dense_matrix,
    dense_vector,
    validate_args=False,
    name=None,
    **kwargs
)

Args:

  • sparse_or_dense_matrix: SparseTensor or Tensor representing a (batch of) matrices.
  • dense_vector: Tensor representing a (batch of) vectors, with the same batch shape as sparse_or_dense_matrix. The shape must be compatible with the shape of sparse_or_dense_matrix and kwargs.
  • validate_args: When True, additional assertions might be embedded in the graph. Default value: False (i.e., no graph assertions are added).
  • name: Python str prefixed to ops created by this function. Default value: 'sparse_or_dense_matvecmul'.
  • **kwargs: Keyword arguments to tf.sparse_tensor_dense_matmul or tf.matmul.

Returns:

  • product: A dense (batch of) vector-shaped Tensor of the same batch shape and dtype as sparse_or_dense_matrix and dense_vector.