I am trying to optimize the function f (g (k), h (k)) according to a list of parameters indicated by the letter k. And there is a restriction on the inequality over some k and a restriction on the equality g (k) . I tried both scipy optimize.minimize and optimize.fmin_slsqp, however, what it seems to him is to change the values of all the parameters in the list k one by one, and then do a few iterations, and then just exit the process with Singular matrix C in LSQ subproblem (Exit mode 6). I have no idea why it is not working. Please note: I do not provide a gradient of constraints, in fact I have a constraint on g (k) .imag = 0 (which is non-differentiable). I tried to remove this and it no longer gives me an error. Can someone explain this behavior.
Thank!
source
share