[molpro-user] Expanding scf/ks procedure

[Jutho Haegeman]

First attempt was rather buggy, and some things didn't work. Here is  
a code that seems to work:

{matrop
! compute S2

diag,US2,eigS2,S2

add,SMH2,+1,S2,-1,S2 ! trick to optain zero matrix

save,US2,2501.2,ORBITALS
save,eigS2,2502.2,SQUARE
save,SMH2,2503.2,SQUARE
}

do j=1,10 ! fixed basissize
{matrop

load,US2,ORBITALS,2501.2
load,eigS2,SQUARE,2502.2
load,SMH2,SQUARE,2503.2

elem,sj,eigS2,j,1

oprd,TMP,US2,j+0.1,j+0.1,1/sqrt(sj)

add,SMH2,SMH2,TMP   ! direct adding in oprd, as described in manual,  
doesn't seem to work
save,SMH2,2503.2,SQUARE
}
enddo

I've checked this procedure by manually calculating S^-1/2 and  
comparing it with the standard SMH-matrix.  So the only remaining  
question is how to determine the basissize, or, alternatively, the  
number of irreps and the number of basisfunctions in a certain irrep.

Thanks for any help with that,

Jutho