Derivatives

The first partial derivatives of the problem functions are computed through "finite differencing," which involves perturbing the current values of the decision variables, observing how the problem functions change, and performing a "rise over run" calculation. With Forward differencing (the default choice), the point from the previous iteration (where the problem function values are already known) is used in conjunction with the current point. This reduces the recalculation time required for finite differencing, which can account for up to half of the total solution time. Central differencing relies only on the current point and perturbs the decision variables in opposite directions from that point. Although this involves more recalculation time, it may result in a better choice of search direction when the derivatives are rapidly changing, and hence fewer total iterations.