optimizationnonlinear-optimizationnon-convexneos-server

Cannot evaluate constraint(s) at starting point

I am using MINLP with NEOS solver, my problem is non-convex and I get this 

Cannot evaluate constraint(s) at starting point

what does this mean and how I can fix it?

Solution

  • The meaning is rather obvious. The functions in your constraints cannot be evaluated at the starting point. If the starting point is zero and you have log(x-0.5) in the model you may get a message like this.

    The fix is to provide better bounds and a better initial point. How to specify this depends on how you created the model and what modeling tools were used.