hi all, i have a few questions that are more about understanding, then finding a solution: i simplified an existing code. it is a maximazition model with many non-linear equations. now as i run the adjusted model, the following warning shows up:

"** Warning ** The number of nonlinear derivatives equal to zero

in the initial point is large (= 21 percent).

A better initial point will probably help the

optimization."

the warning doesn't show up in the original code. otherwise the adjusted code works fine and the results make sense. my questions:

what does the warning mean exactly? how do starting values influence the derivates?

what might be the implication of this "suboptimal' initial point?

in the original code the model is run 3 times in a row, so i do the same in the adjusted code, for the second and third solve, the warning message doesn't appear:

solve model maximizing utility using nlp;

solve model maximizing utility using nlp;

solve model maximizing utility using nlp;

what is the reason for doing this? what does running the model more then once change/is the second run influenced by the first?

why does the warning disappear?

thank you for clarifying, i didn't find anything about this warning on the gams website.