thirdOrderLossFunction

open override fun thirdOrderLossFunction(x: Double): Double(source)

Third order loss function G3(x) = (1/6)Emax(X-x,0)*max(X-x-1,0)*max(X-x-2,0).

All three factors are clamped. An empirical support can place a point strictly between x and x+2, where a later factor turns negative while an earlier one has not, and a product of clamped non-negative quantities cannot be negative.

Requires a whole-number support, for the reason given on secondOrderLossFunction: the accumulation that defines this order runs over the integers, and a support off them has none.

Throws

if any support point is not a whole number