$MCSCF

$MCSCF group (optional for -MCSCF-)

This group controls the MCSCF orbital optimization step. The FULLNR method almost always exhibits good convergence, but is more time consuming than the FOCAS or SOSCF methods. All benefit greatly from well chosen starting orbitals, and your having carefully read the section on MCSCF in Chapter Four.

--- the next three choose the optimization strategy ---

FOCAS
= a flag to select a method, applicable only to FORS type MCSCF functions, with first order convergence rate. (default=.FALSE.)

SOSCF
= a flag selecting an approximately second order convergence method, which is applicable only to FORS type MCSCF functions. (default=.TRUE.)

FULLNR
= a flag selecting a second order method, with an exactly computed orbital hessian. The method is usable for FORS or non-FORS functions, although convergence is harder with the latter. The memory demands of this step are high, as is the computational load. (default=.FALSE.)

--- the next apply to all three convergence methods ---

MAXIT
= Maximum number of iterations (default=100 for FOCAS, 60 for SOSCF, 30 for FULLNR)

MICIT
= Maximum number of microiterations within a single MCSCF iteration. (default=5 for FOCAS or SOSCF, or 1 for FULLNR))

ACURCY
= the major convergence criterion, the maximum permissible asymmetry in the Lagrangian matrix. (default=1.0E-05)

ENGTOL
= a secondary convergence criterion, the run is considered converged when the energy change is smaller than this value. (default=1.0E-10)

DAMP
= damping factor, this is adjusted by the program as necessary. (default=0.0)

NWORD
= The maximum memory to be used, the default is to use all available memory. (default=0)

CANONC
= a flag to cause formation of the closed shell Fock operator, and generation of canonical core orbitals. This will order the MCC core by their orbital energies. (default=.TRUE.)

EKT
= a flag to cause generation of extended Koopmans' theorem orbitals and energies. (Default=.FALSE.)
For this option, see R.C.Morrison and G.Liu, J.Comput.Chem., 13, 1004-1010 (1992). Note that the process generates non-orthogonal orbitals, as well as physically unrealistic energies for the weakly occupied MCSCF orbitals. The method is meant to produce a good value for the first I.P.

NPUNCH
= MCSCF punch option (analogous to $SCF NPUNCH)

0 do not punch out the final orbitals
1 punch out the occupied orbitals
2 punch out occupied and virtual orbitals

The default is NPUNCH = 2.

NPFLG
= an array of debug print control. This is analagous to the same variable in $CIINP. Elements 1,2,3,4,6,8 make sense, the latter controls debugging the orbital optimization.

--- the next refers to SOSCF optimizations ---

NOFO
= set to 1 to skip use of FOCAS for one iteration during SOSCF. This is a testing parameter, at present NOFO defaults to 0 to do one FOCAS iter.

--- the next three refer to FOCAS optimizations ---

CASDII
= threshold to start DIIS (default=0.05)

CASHFT
= level shift value (default=1.0)

NRMCAS
= renormalization flag, 1 means do Fock matrix renormalization, 0 skips (default=1)

--- all remaining input applies only to FULLNR ---

FORS
= a flag to specify that the MCSCF function is of the Full Optimized Reaction Space type, which is sometimes known as CAS-SCF. The default is to optimize act-act rotations. These must be optimized if FORS was not set in $DRT, and often convergence is better with them included even for FORS wavefunctions. (default=.FALSE.)

METHOD
= DM2 selects a density driven construction of the Newton-Raphson matrices. (default).

= TEI selects 2e- integral driven NR construction.

= FORMULA selects formula driven NR construction. (the FORMULA method is now an inactive option) See the 'further information' section for more details concerning these methods.

LINSER
= a flag to activate a method similar to direct minimization of SCF. The method is used if the energy rises between iterations. It may in some circumstances increase the chance of converging excited states. (default=.FALSE.)

FCORE
= a flag to freeze optimization of the MCC core orbitals, which is useful in preparation for RUNTYP=TRANSITN jobs. Setting this flag will automatically force CANONC false. This option is incompatible with gradients, so can only be used with RUNTYP=ENERGY. (default=.FALSE.)

--- the next two are seldom used ---

NORB
= the number of orbitals to be included in the optimization, the default is to optimize with respect to the entire basis. This option is incompatible with gradients, so can only be used with RUNTYP=ENERGY. (default=number of AOs given in $DATA).

NOROT
= an array of up to 250 pairs of orbital rotations to be omitted from the NR optimization process. The program automatically deletes all core-core rotations, all act-act rotations if FORS=.T., and all core-act and core-virt rotations if FCORE=.T. Additional rotations are input as I1,J1,I2,J2... to exclude rotations between orbital I running from 1 to NORB, and J running up to the smaller of I or NVAL in $TRANS.

--- The next 3 pertain only to METHOD=FORMULA ---

FMLFIL
= a flag specifying the formula tape has been saved for reuse in this run. The default is to regenerate this file on the 1st iteration at the 1st geometry run. (default=.FALSE.)

LDAR
= length of each direct access record during the sort of the formula tape. (default is machine dependent)

NDAR
= number of direct access records. The value used will be the larger of the program's guess of the actual number, and the input value (default=0).


Back to list of input groups...