21.4.17 Miscellaneous options

OPTION,code1=value,code2=value,$\ldots$

Can be used to specify program parameters and options. If no codes and values are specified, active values are displayed. The equal signs may be omitted. The following codes are allowed (max 7 per card):

NSTATE:

see state card

NSTATI:

number of states calculated in internal CI

NSTATR:

see refstat card

NCEPA:

see CEPA card

NOKOP:

if nonzero, skip integral transformation

ITRDM:

if .ge. 0 transition moments are calculated

ITRANS:

if nonzero, perform full integral transformation (not yet implemented)

IDIP:

Print dipole moments from iteration number value

REFOPT:

if nonzero, optimize reference coefficients; otherwise extract reference coefficients from internal CI

IAVDEN:

average HII and HSS denominators over spin couplings if nonzero

IDELCG:

if.ne.0 then destroy files icfil,igfil at end

IREST:

if nonzero, restart

NATORB:

if nonzero, natural orbitals are calculated and printed. The number of printed external orbitals per symmetry is min(natorb,2)

WFNAT:

if nonzero, natural orbitals are saved to this record

IPUNRF:

if nonzero, punch coefficients of reference configurations

NPUPD:

if nonzero, update pairs in nonorthogonal basis, otherwise in orthogonal basis.

MAXIT:

see maxiter card

MAXITI:

see maxiter card

MAXDAV:

see maxdav card

MAXVI:

see maxdav card

NOSING:

see nosing card

NOPAIR:

see nopair card

MXSHRF:

see select card

IKCPS=0:

In CIKEXT, only K(CP) is calculated; this option taken when and only when no singles.

IKCPS=1:

only K(CP') is calculated. Implies that modified coupling coefficients are used.

IKCPS=2:

K(CP) and K(CP') are calculated. Default is IKCPS=2 except when single reference configuration, when IKCPS=1.

IOPTGM:

Option for density matrix routines.

IOPTGM=0:

all quantities in density matrix routines are recalculated for each intermediate symmetry (max. CPU, min. core).

IOPTGM=1:

quantities precalculated and stored on disk (max. I/O, min. core).

IOPTGM=2:

quantities precalculated and kept in core (min. CPU, max. core).

IOPTOR:

If nonzero, calculate intermediate orbitals for each pair. Might improve convergence in some cases, in particular if localized orbitals are used.