Posted by Jellby on March 21, 2007 at 11:09:22:
In Reply to: Hessian calculation posted by Jellby on March 12, 2007 at 18:27:44:
: While doing some tests I found that the Hessian given by MCKINLEY is different from the Hessian I get, for example, with Gaussian98. I made a calculation for H2 (SCF/STO-3G), and I got the same energy and gradient in both programs, but the Hessian is not the same.
Well, I've just calculated the numerical Hessian manually (by calculating energies and gradients at displaced geometries), and I'm getting with both Molcas and Gaussian the same numerical Hessian, which is identical (allowing for precision) to the analytical Hessian given by Gaussian, but different from the one given by Molcas.
It seems there is either a bug in MCKINLEY or some component or conversion that I'm missing.