probe_derivatives_kind
======================

.. py:function:: t3toolbox.backend.fitting.probe_derivatives_kind(order, weight = None, chunk_size = 100)

   .. code-block:: python

      def probe_derivatives_kind(
              order:      int,
              weight:     typ.Optional[typ.Any] = None,   # residual weight ω[mode,order], (d,order+1) broadcast; None = 1
              chunk_size: typ.Optional[int] = 100,        # W-chunk size for the 𝒥ᵀ gradient assembly (docs/chunking.md)
      ) -> SamplingKind:                                  # sample = (ww, pp); data = list of d, (order+1)+W+(Ni,)


   The **probe-derivatives** sampling kind: vector-valued (one free mode per probe), so the residual
   / output is a list of ``d`` arrays. ``sample = (ww, pp)``. Probe has both a mode and an order axis, so
   ``weight`` is the full ``ω[mode, order]`` matrix ``(d, order+1)`` (a row ``(order+1,)`` = per-order, a
   column ``(d, 1)`` = per-mode, a matrix = both).