Hi,
I'm trying to compare some of the benchmark studies done with NWCHEM's rt_tddft module against my code. I'm referring to the following paper to understand the numerical implementation:
K. Lopata, N. Govind,"Modeling Fast Electron Dynamics with Real-Time Time-Dependent Density Functional Theory: Application to Small Molecules and Chromophores", J. Chem. Theory Comput., 7, 1344 (2011).
For the Magnus propagator, the paper mentions a predictor-corrector scheme to evaluate the Fock matrix at a future time. The predictor step involves a linear extrapolation using Fock matrices at two previous timesteps. I was wondering how is the predictor step performed for the first two time-steps.
Thanks,
Bikash
|