tf.contrib.distributions.fill_triangular_inverse

Creates a vector from a (batch of) triangular matrix.

tf.contrib.distributions.fill_triangular_inverse(
    x, upper=False, name=None
)

The vector is created from the lower-triangular or upper-triangular portion depending on the value of the parameter upper.

If x.shape is [b1, b2, ..., bB, n, n] then the output shape is [b1, b2, ..., bB, d] where d = n (n + 1) / 2.

Example:

fill_triangular_inverse(
  [[4, 0, 0],
   [6, 5, 0],
   [3, 2, 1]])

# ==> [1, 2, 3, 4, 5, 6]

fill_triangular_inverse(
  [[1, 2, 3],
   [0, 5, 6],
   [0, 0, 4]], upper=True)

# ==> [1, 2, 3, 4, 5, 6]

Args
x	Tensor representing lower (or upper) triangular elements.
upper	Python bool representing whether output matrix should be upper triangular (True) or lower triangular (False, default).
name	Python str. The name to give this op.

Returns
flat_tril	(Batch of) vector-shaped Tensor representing vectorized lower (or upper) triangular elements from x.