Fig. 1: Truncation errors in finite difference operators.
From: Adaptive finite differencing in high accuracy electronic structure calculations
a, b Absolute value of the finite difference truncation error of plane waves with different G vectors, up to half of the Nyquist frequency Gmax. The 8th-order AFD operator uses a mixing value of M = 0.33. Part b is an enlargement of the low-G part of a. The vertical axes units are 1/h2 (see Eq. 7).
