I’m having trouble unfolding the G3 matrix generated by Dynare (oo_.dr.g_3) in the 3rd order perturbation solution.
The G3 matrix must be unfolded in order to simulate from the policy functions simulations outside Dynare using the perturbation results.

This problem has afflicted other users that eventually managed to unfold the G2 matrix, see e.g. the discussion in:

viewtopic.php?f=1&t=2851#p6548

QUESTION: I wonder if anyone has ever managed to unfold the G3 matrix? (and how?)