ut3_inner#

t3toolbox.backend.ut3_linalg.ut3_inner(x, y, use_orthogonalization=True)#
def ut3_inner(
        x:  UT3Data,                                  # (tucker_supercore, tt_supercore, shape, masks)
        y:  UT3Data,                                  # same physical shape; ranks/masks/padding may differ
        use_orthogonalization: bool = True,           # True (default): orthogonalize both first (stable)
) -> NDArray:                                         # HS inner product <T(x), T(y)>, shape=stack_shape

Hilbert-Schmidt inner product of two uniform Tucker tensor trains – the twin of the ragged t3_inner_product. See ut3_norm() for the orthogonalization / autodiff story.

Parameters:

  • x (UT3Data)

  • y (UT3Data)

  • use_orthogonalization (bool)

Return type:

NDArray