Shuffled Complex Evolution

[NICOLÁS ANDRÉS VÁSQUEZ]

Hello everyone,

I am calibrating the VIC model using the Shuffled Complex Evolution algorithm and I have a doubt. When the calibration process is completed, the "optimum" parameter set is not the maximum value that OSTRICH found during the process. I attached a figure with the objective functions during the calibration process (black dot) and the "optimum" value according to OSTRICH (red line). Is this ok? Am I misunderstanding something?

Many thanks,

Nico Vásquez

[Mehmet Cüneyd Demirel]

Hi Nico,

Do you use the result in OstNonDomSolutions0.txt?

I believe that you removed "1-" from the KGE formula or a similar manipulation in ostin.txt to let the optimizer find fmin as zero.

Cheers,

Cüneyd