Safely Stop a Pareto Search

9 次查看（过去 30 天）

显示更早的评论

Ayman Al-Sukhon 2020-5-14

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/525500-safely-stop-a-pareto-search

评论： Ameer Hamza 2020-5-14

采纳的回答： Ameer Hamza

在 MATLAB Online 中打开

Hi,

I am running a Pareto search with an expensive cost function. I limited the time to 39600 s using the 'MaxTime' option however it has been more than 12 hours (43200 s):

fun = @CostFunctionOptim_Pareto;
options = optimoptions('paretosearch','ParetoSetSize',200,'InitialPoints',x_seed,'PlotFcn','psplotparetof','MaxTime',39600);
[x,fval,exitflag,output] = paretosearch(fun,4,Aineq,bineq,Aeq,beq,lb,ub,nonlcon,options);

I don't know why it's still going, but I don't want to stop it for fear that it will erase all the progress it has so far (I am happy with the results it has so far and don't need more points in the Pareto front). Is there any way to safely stop the optimzer and still get the output values?

Thanks

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

采纳的回答

Ameer Hamza 2020-5-14

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/525500-safely-stop-a-pareto-search#answer_432517

I hope you haven't interrupted the optimization already. The safest way I can think of is to add a breakpoint inside your objective function. The optimizer will stop at that point, and you will have access to the current values of the optimization variables. You can then use save() to save the values of optimization variables in a .mat file, and then you can interrupt the paretosearch. Now you have the latest optimal values of the variable in a mat file, load that file and use those those values.