problems with multiple solve statements in AMPL

[Hossein Haghighat]

Hi,
I solve my problem for several scenarios in a loop. Each time I solve the problem individually, I find a solution but when I solve the problem in a loop I get another solution (and sometimes the same model that is solved does not get solved in the loop!!).
I figured out that this happen because AMPL starts the next optimization process form the previous solution and if the previous solution is not a good starting point difficulties arise. Am I right?
please advise me how to prevent AMPL from starting the next solve command from the previous answer. is there any "option" in AMPL to avoid this? 
 

---

Explore the seven wonders of the world Learn more!

[Robert Fourer]

Set the following:

 

   option reset_initial_guesses 1;

 

 

---

--~--~---------~--~----~------------~-------~--~----~