lets assume a function
f(x,y) = z
Now I want to choose x so that the output of f matches real data, and y decreases in equidistant steps to
