[molpro-user] A problem in the rotation of orbitals

[Kirk Peterson]

Dear Rana,

which version of Molpro are you using?  Does the output from multi not print that these two orbitals have been
rotated?  Generally rotation by 0 degs as you've specified is equivalent to rotation by 90, which is to swap.  Of course
even if you swap these two initial orbitals, there is no reason why in the course of the MCSCF iterations the character
of these orbitals couldn't switch back again if it resulted in a lower energy.

On another note, the record you're specifying in your orbital directive seems incorrect to me.  These should have a
syntax something like  XXXX.2, where XXXX is the record number (typical value around 2100 to 4100) and the .2 specifies
to use Molpro temporary file 2 (which you are saving to wf0rotta2a1)

Also if the multi program requires more than 10 iterations to converge it almost always indicates a problem with the 
definition of the active space.

best,

Kirk


On Feb 14, 2012, at 12:40 AM, Rana Zahdeh wrote:

> Dear MOLPRO users,
>  
> I face a problem in the convergence of the MCSCF wave function, so I'm trying to rotate two orbitals,
> Wishing that this may help in the convergence of the MCSCF. Below is my input file, I tried to rotate the 
> orbital 59.2 (which is from the active space) with 48.2 (which is not from the active space), but when I 
> draw the orbitals with molden I found that the orbital's graph of 59.2 is not exchanged with the orbital's 
> graph of 48.2. Why this exchange didn't take place? is there a problem with my input file?
>  
> memory,150,m
> gprint,orbital,civector
> file,1,/scratch/ranaobaid/wfu/int0rottta2a1
> file,2,/scratch/ranaobaid/wfu/wf0rottta2a1
> basis=vdz             !define basis
> fir=0
> fi=fir
> s=sin(fi)
> c=cos(fi)
> dfi=5e-3
> ffir=0    !define bond distances
> ffi=ffir
> ss=sin(ffi)
> cc=cos(ffi)
> dffi=5e-3
> geomtyp=xyz
> include cpoxoequi.inp
> hf
> 
> {multi;occ,27,8,19,5;closed,27,3,18,1;wf,110,1;state,2;weight,0,1;orbital,245410;start,245410;rotate,59.2,48.2,0;maxiter,80,80}
> put,molden,mcscfrot1.molden
> 
> Note, the above input file is converged when I used the multi command as follows:
>  
> {multi;occ,27,8,19,5;closed,27,3,18,1;wf,110,1;orbital,245410;maxiter,80,80}
> 
> Will you please help me to know what the problem is?
>  
> Looking forward to hearing from you....
>  
> Thanks in advance,
> Rana.
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20120214/668a1a4a/attachment.html>