Helper class for computing samples along a Gaussian distribution curve.
The curve is defined by this function (and its inverse):
y = e ^ (((x - expected) / deviation) ^ 2 / -2) * scale x = (ln(y / scale) * -2) ^ 0.5 * deviation + expected
expected is the expected value of the distribution,
deviation is the standard deviation and
scale is a factor chosen so that Compute returns the specified maximum value when passing
The serialization helper object for values of Gauss.