Orbitals in Molpro

[Kirk Peterson]

I agree - at least G98 does not support general contractions, so
it's doing something different here (I believe using the method
suggested by Davidson). Since you're running this with a contracted
basis set on the atom, the HF 1s orbital of hydrogen should
have a coefficient of exactly 1.0 for the 1s contraction (as molpro gives).

best wishes,

Kirk

> 
> the scheme of function contraction is different. Use the 
> input stream format of the basis set and the results will coincide.
> wishes
> Elmer
> 
> On Wed, 13 Mar 2002 sasha at okra.cchem.berkeley.edu wrote:
> 
> > Dear Molpro users,
> > 	I'm attempting to extract a trial wave function from Molpro to be
> > used in a set of QMC calculations.  As a first test I am using the  
> > hydrogen atom with HF and the cc-pVDZ basis, turning off the symmetry and
> > using cartesian basis functions.  Compare the resulting orbital
> > coefficients from those obtained using GAMESS and Gaussian, the
> > coefficients obtained in MOLPRO are rather different. I've enclosed the
> > results: 
> > 
> > MOLPRO:     1.000001   -0.000001    0.000000     0.000000   0.000000
> > GAMESS/G98  0.587634722 0.501238983 0.00000000   0.00000000 0.00000000
> > 
> > Can any one explain why this difference occurs? Is there a way to
> > transfrom from one set to another? Any information is greatly appreciated.
> > regards
> > Alex
> > 
> > 
> > 
> 
> 

----- End of forwarded message from Elmer Valderrama -----

-- 

------------------------------------------------------------------------------
Kirk A. Peterson
Associate Professor              Affiliate Senior Research Scientist
Department of Chemistry     &    Theory, Modeling, and Simulation
2710 University Dr.              Environmental Molecular Sciences Laboratory
Washington State University      Pacific Northwest National Laboratory
Richland, WA 99352               P.O. Box 999, Mail Stop K8-91

Office: (509) 376-2023, (509) 372-7282
Fax:    (509) 376-0420
kirk.peterson at pnl.gov
http://www.tricity.wsu.edu/~kipeters/
------------------------------------------------------------------------------