Abstract
The authors pointed out the disagreement between their calculation of dynamical correlations and that obtained using the exact result as in [2] can be attributed to two of following reasons: (i) The linear ansatz overestimates the higher coefficients and, given the nonoscillatory growth rate of coefficients with respect to n, suppresses the oscillation [1] in a0(t) thus damping it out faster. (ii) The inverse Laplace transform calculation is typically accurate to 10-4 or so unless a higher accuracy is specifically being sought [1].
- Received 15 October 1993
DOI:https://doi.org/10.1103/PhysRevLett.72.3287
©1994 American Physical Society