I have recently been running some analyses using the minimax
optimization criteria, and thus the simplex solver. During some
analyses, the MinMaxNRSimplex solver would endlessly cycle. When using
the MinMaxSimplex solver, the analysis could be completed without
stalling. I have read that the MinMaxSimplex solver is more robust for
some situations.
With the robustness in mind, can the solvers output anything regarding
whether or not the optimization routine actually terminated at an
optimal point? For example, when muscle forces are desired, can the
Kuhn-Tucker conditions (or information for computing them) be output?
Unfortunately there is no way to get this info out from the system, sorry,
but in the study it is possible to export the equilibrium equations to a
text file.
I have never had any reason to think that the recruitment stopped in a non
optimal point, this should be seen as jumps on the muscles activity
envelope.
I have recently been running some analyses using the minimax
optimization criteria, and thus the simplex solver. During some
analyses, the MinMaxNRSimplex solver would endlessly cycle. When using
the MinMaxSimplex solver, the analysis could be completed without
stalling. I have read that the MinMaxSimplex solver is more robust for
some situations.
With the robustness in mind, can the solvers output anything regarding
whether or not the optimization routine actually terminated at an
optimal point? For example, when muscle forces are desired, can the
Kuhn-Tucker conditions (or information for computing them) be output?
Thanks,
Chris
[Non-text portions of this message have been removed]