Instead of doing this job I thought: doing optimization + forward over a long period and starting the tests on the future of the best
results,for example 100, with overlapping graphic display of the 100 results. In this way there would be the performance lines one
above the other and an immediate evaluation of the parameters that could work.
Yes, in fact, I write this because it is a month that I select the results after calculations of hours over 3 months, 6 months and more, ticks, real ticks etc .. of the optimization and in demo they suck from the second day