29f7321b64ada5e66ff4ba8e4c9d0a44fed241f4,t3f/initializers.py,,random_tensor_batch,#Any#Any#Any#Any#Any#,244

Before Change

    tt_rank = np.append(tt_rank, 1)
  tt_rank = tt_rank.astype(int)
  // TODO: variable (name?) scope.
  tt_cores = [None] * num_dims
  for i in range(num_dims):
    curr_core_shape = (batch_size, tt_rank[i], shape[i], tt_rank[i + 1])
    tt_cores[i] = tf.random_normal(curr_core_shape, mean=mean, stddev=stddev)

  return TensorTrainBatch(tt_cores, shape, tt_rank, batch_size)


def random_matrix(shape, tt_rank=2, mean=0., stddev=1.):
  Generate a random TT-matrix of given shape.

After Change

      raise ValueError("all elements in `tt_rank` should be integers")
  if not isinstance(batch_size, (int, np.integer)):
    raise ValueError("`batch_size` should be integer")
  num_dims = shape.size
  if tt_rank.size == 1:
    tt_rank = tt_rank * np.ones(num_dims - 1)
    tt_rank = np.insert(tt_rank, 0, 1)
    tt_rank = np.append(tt_rank, 1)
  tt_rank = tt_rank.astype(int)

  cr_exponent = -1.0 / (2 * num_dims)
  var = np.prod(tt_rank ** cr_exponent)
  cr_stddev = stddev ** (1.0 / num_dims) * var
  tt = tensor_batch_with_random_cores(shape, tt_rank=tt_rank, stddev=cr_stddev,
                                      batch_size=batch_size)

  if np.abs(mean) < 1e-8:
    return tt
  else:
    raise NotImplementedError("non-zero mean is not supported yet")


def random_matrix(shape, tt_rank=2, mean=0., stddev=1.):
  Generate a random TT-matrix of the given shape with given mean and stddev.

  Entries of the generated matrix (in the full format) will be iid and satisfy

Instances