GPO minimizer not working as expected

**Description of the error**
I've been trying to optimise the FoM using GPO, however, it just gets stuck and does nothing useful.

**Additional details**
Here I run the examples/argon.py script on the chi/why-wont-argon-just-work branch which has the argon.py script I am using. I fix the scaling factor to 18.5 which gives a small values of the FoM for sigma=3.3620 and epsilon=1.0200.

The FoM calculation at the same point is somewhat stable. Here are five runs from MDMC.

FoM epsilon (#2) sigma (#3)
0.8958 1.02 3.362
1.564 1.02 3.362
0.9887 1.02 3.362
0.8737 1.02 3.362
1.382 1.02 3.362

So it should optimise to this point or another similar to this one. However, it just seems to get stuck and not do anything particularly useful. Here are the best results after 595 iterations.

note that the lowest value was generated during the initial points not during the actual optimisation.

Here is another run, but I use the ln(FoM) with 441 iterations.