Hi Adel,
I'd say you need to negate the return value, according to the user manual,
nlopt only takes nonlinear constraints of the form g(x) <= 0.
so:
... InequalityConstraint(...)
{
return( -1.0*(x[2]*x[2]+x[1]) )
}
also, you should consider passing a tolerance when you add the constraint...
that's also mentioned in the manual :)
-Sam
________________________________
From: NLopt-discuss [[email protected]] on behalf of Adel
M [[email protected]]
Sent: Monday, December 09, 2013 4:30 PM
To: [email protected]
Subject: [NLopt-discuss] COBYLA with constraints
Hi,
I have a problem when adding an inequality constraint in the form of
X[2]*X[2]+X[1]>=0 that seems not to taken into account by the optimizer
My C++ code looks like:
static double InequalityConstraint(const std::vector<double>& x,
std::vector<double> &grad, void *data)
{
return (x[2]*x[2]+x[1]);
}
and
opt1.add_inequality_constraint(InequalityConstraint,NULL);
(opt1 is my nlopt::opt )
I cannot figure out why this constraint is not taken into account!
Best Regards,
Adel
_______________________________________________
NLopt-discuss mailing list
[email protected]
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss
