ut3_apply_derivatives

t3toolbox.backend.ut3_sampling.ut3_apply_derivatives(ww, pp, data, order)

def ut3_apply_derivatives(
        ww:    typ.Sequence[NDArray],  # apply vectors X,        len=d, ith elm_shape=W+(Ni,)
        pp:    typ.Sequence[NDArray],  # perturbation vectors P, len=d, ith elm_shape=W+(Ni,)
        data:  UT3Data,
        order: int,                    # highest derivative order
) -> NDArray:                          # shape=(order+1,)+W+stack_shape

Symmetric all-modes apply derivatives of a uniform Tucker tensor train (shares sampling_derivatives.t3_apply_derivatives; a scalar jet per stack element). ww/pp packed to N.

Parameters:

• ww (t3toolbox.backend.common.typ.Sequence[NDArray])

• pp (t3toolbox.backend.common.typ.Sequence[NDArray])

• data (UT3Data)

• order (int)

Return type:

NDArray