Next: Linear reponse calculation: *LINEAR
Up: **RESPONSE
directives
Previous: End of input: *END
General-purpose directives are given in the **RESPONSE
section.
- .NOAVDI
Do not use Fock type decoupling of the two-electron density matrix.
Add
instead of
to
approximate
orbital diagonal.
- .INPTES
Input test. For debugging purposes only. The program stops after the
input section.
- .HIRPA
Invoke the higher RPA approximation for the calculation of linear
response properties. This approximation is identical to that of McKoy
and coworkers [85, 86]. The requirements to the
preceding wave function
calculation is the same as for the .SOPPA
keyword.
- .MAXPHP
READ *, MAXPHP
Change the maximum dimension of
subspace. Default is 100.
PHP is a subblock of the CI matrix which is calculated explicitly
in order to obtain improved CI trial vectors compared to the
straight Davidson algorithm[87]. The configurations
corresponding to
the lowest diagonal elements are selected, unless .PHPRESIDUAL is
specified. MAXPHP is the maximum dimension of PHP, the
actual dimension will be less if MAXPHP will split degenerate configurations.
- .MAXRM
READ *, MAXRM
Change the maximum dimension of the reduced space. Default is 200.
When solving a linear system of equations or an eigenvalue equation,
the reduced space is increased by 1 in each iteration. For single root
calculations this should exceed the number of iterations required.
- .NODOIT
Turns off direct one-index transformation [88].
In this way all one-index
transformed integrals are stored on disk.
- .NOITRA
No integral transformation. Normally the two-electron integrals are
transformed in the beginning of a response calculation. In some cases
this is not desirable, e.g., if the response part is only used for
calculating average values of operator, or if the transformed two-electron
integrals already exist from a previous response calculation.
- .S0MIX
Sum rule is calculated in mixed representation, that is, calculate
provided that dipole length and
velocity integrals are available on the property integral file
(calculated with **HERMIT
options .DIPLEN
and .DIPVEL
).
The calculated quantity gives a measure of the quality of the basis
set .
- .OPTORB
Orbital trial vectors are calculated with optimal orbital
trial vector
algorithm [7].
- .ORBSFT
READ *, ORBSFT
Change the amount for shifting the orbital diagonal of the MCSCF Hessian.
May be used if there is a large number of negative eigenvalues.
Default is
.
- .ORBSPC
Calculation with only orbital operators.
- .PHPRES
Select configurations for PHP matrix based on largest residual
rather than lowest diagonal elements.
- .PROPAV
READ '(A)', LABEL
Property average . The average value of an
operator is calculated.
The line following this option must contain the
label of the operator given in the integral property file.
(See section
.)
- .QRREST
Restart quadratic response calculation.
- .RSPCI
Configuration interaction
response calculation.
- .SOPPA
Required, no defaults.
The SOPPA flag requires that
the preceding SIRIUS calculation has generated the MP2 correlation
coefficients and written them to disk. (Set .RUN RESPONSE
in **DALTON
input as well as .HF
and .MP2
in **WAVE FUNCTION
). See
example input in Chapter
.
- .SOPW4
Calculate explicitly the W4 term described by Odderhede et
al. [89]. This term is already included in the normal
SOPPA result, and used mostly for comparing to older
calculations. Note that this keyword requires that .SOPPA
is set.
- .TRPFLG
Triplet flag. This option is set whenever triplet
(spin-dependent)
operators must be used in a response calculation
[90, 91]
Next: Linear reponse calculation: *LINEAR
Up: **RESPONSE
directives
Previous: End of input: *END
Kenneth Ruud
Sat Apr 5 10:26:29 MET DST 1997