fv_absorb_weights
=================

.. py:function:: t3toolbox.backend.fv_operations.fv_absorb_weights(variations, weights)

   .. code-block:: python

      def fv_absorb_weights(
              variations: typ.Tuple[typ.Sequence[NDArray], typ.Sequence[NDArray]],  # (tucker_variations V, tt_variations H)
              weights:    typ.Tuple[typ.Sequence[NDArray], typ.Sequence[NDArray],
                                    typ.Sequence[NDArray], typ.Sequence[NDArray]],   # (up, down, left, right), each len=d
      ) -> typ.Tuple[
          typ.Tuple[NDArray, ...],  # weighted tucker_variations, elm_shape=stack+(nDi, Ni)
          typ.Tuple[NDArray, ...],  # weighted tt_variations,     elm_shape=stack+(rLi, nUi, rRi)
      ]:


   Absorb the four-family metric weights into the VARIATION cores (the tangent metric on coordinates,
   Approach-1 / metric-on-variations): ``down`` -> V's ``nD`` leg; ``up``/``left``/``right`` -> H's
   ``nU``/``rL``/``rR`` legs. The frame is left orthonormal and untouched. ``corewise_stack_norm`` of the
   result is the weighted (Grasedyck-Kramer) tangent norm. All families are len=d (one per variation core);
   single leading ``'...'`` -- the weights share the variations' ``K+C`` stack.