[molpro-user] Geometry optimization method

H. -J. Werner werner at theochem.uni-stuttgart.de
Mon Oct 24 18:44:53 BST 2005


Dear Nicholas,
I think the qsd method is the default since some time since it
usually converges faster than rf, and it appears that the
documentation is incorrect. But method,rf should use the
Rational Function method.
Joachim Werner

On Mo, 24 Okt 2005, Nicholas Marshall wrote:

>Dear Molpro users,
>
>The manual states that geometry optimizations use the Rational Function 
>method by default. However, several times when running optimizations I 
>have seen the line
>
> Quadratic Steepest Descent - Minimum Search
>
>appear in my output. I have not set the "method" keyword at all, 
>relying on the default. This seems to indicate that the program is 
>using the Sun and Ruedenberg steepest-descent method (method, qsd) 
>which would not be an appropriate method for this application. In the 
>test case I've been looking at, setting method,rf explicitly makes the 
>steepest-descent line go away. Why is it doing this?
>
>Thanks,
>
>Nicholas Marshall
>
>------------------------------------
>relevant part of input: (the basis is just so test jobs run rapidly)
>
> basis=STO-3G
>
>   rhf;printorb,1;maxiter,200;
>   ccsd(t);
>   optg,maxit=1;
>   print,gradient=1;
> !  method,rf;

-- 
Prof. Hans-Joachim Werner
Institute for Theoretical Chemistry
University of Stuttgart
Pfaffenwaldring 55
D-70569 Stuttgart, Germany
Tel.: (0049) 711 / 685 4400
Fax.: (0049) 711 / 685 4442
e-mail: werner at theochem.uni-stuttgart.de



More information about the Molpro-user mailing list