(Specific details: To compute the relative error, the tester library dividesthe difference between the actual and expected values by theaverage of the actual and expected values, to remove a slight numericalbias in the computation.)
(Specific details: To compute the relative error, the tester library dividesthe difference between the actual and expected values by theaverage of the actual and expected values, to remove a slight numericalbias in the computation.)