Clarify also files for analysis

[JJ]

Not clear what the generated file is. It looks like all experiments are one behind the other. Also fopt seems to be the optimum. Is that correct? Wouldn't it be better a different file per experiment or benchmark?

[mariosky]

There is a file for each Function-Dimension, each group is for an instance of that function. Each instance has its own optima. This is needed by COCO, the advantage is that we can compare our results, with more than fifty algorithms from past workshops.