Posted by Jellby on March 23, 2007 at 12:13:28:
In Reply to: Re: Hessian calculation posted by Roland Lindh on March 23, 2007 at 10:27:50:
: there is a small misunderstanding here I see. The Hessian has two contributions. The first one is computed by Mckinley, the tracing of zeroth other densities with 2nd order derivatives. The second part is due to the wavefunction response to the perturbation. It is the MCLR code which computes this (in your input you have disabled the automatic execution of the MCLR module and you have not added an explicit request for it to be computed!!!). Hence, the MCLR module adds to the Hessians which are first order densities traced with first order integral derivatives.
I see, thanks... But as far as I can see, MCLR does not print this contribution to the Hessian, it only prints the frequencies and normal modes. I guess I could get the Hessian from that, but is not there an easier way to get the full Hessian, just the second derivatives of the energy with respect to nuclear displacements?