ccq
is a coupled-cluster (CC) software package capable of running calculations with up to
quadruply excited cluster components. This package includes the standard CCSD, CCSDT,
and CCSDTQ approaches, the active-space CCSDt, CCSDtq, and CCSDTq methods, and the CR-CC(2,3) and CC(t;3)
approaches based on the CC(P ; Q) hierarchy.
A BLAS
library is required to compile this program. It is strongly advised
to use an optimized version, such as OpenBLAS or Intel's MKL.
To compile, just link or copy a configuration file in config/
to
config.mk
. Make sure that it contains the right configuration options (e.g.
compilers, libraries, etc.)
For example:
ln -s config/gfortran.mk config.mk make -j4
The configuration file must include the following information for any calculation type
- froz:
- number of frozen spatial orbitals
- nel:
- number of electrons (occupied spin-orbitals)
- nvir:
- number of virtual orbitals (unoccupied spin-orbitals)
- multiplicity:
- spin multiplicity
- onebody:
- path to the onebody.inp integral file as provided by CC_PACKAGE
- twobody:
- path to the twobody.inp integral file as provided by CC_PACKAGE
- tol:
- energy convergence tolerance. Can be an integer, in which case tol=10^tol. The default value is 1.0E-4.
- shift:
- shift energy used to help converge certain difficult calculations. It's used at the end of each Jacobi iteration and shifts the denominator's value.
- diis_space:
- the size of the DIIS space. Default is 5.
- max_iterations:
- maximum number of iterations before exiting the calculation. Default is 60.
- calc_type:
- select calcualtion type. Currently the supported calculations are CCSD, CCSDT, CCSDTQ, CCSDt, CC(t;3), CADFCIQMC, and DCSD-MC.
- label:
- project label. Will be shown in the output. Useful to help identify the calculation.
- num_threads:
- number of threads to use in the shared-memory parallel sections (default is 1)
- sym_file:
- symmetry file containing the spatial symmetries of the molecular orbitals. The first value of the file is the point group (C1 is 1; C2v, 3; D2h, 4)
- rhf:
- force a closed-shell calculation (default is .false.)
- act_occ:
- number of active occupied spatial orbitals
- act_unocc:
- number of active unoccupied spatial orbitals
- act_ind_t:
- number of active indices in a triply excited cluster
- act_ind_q:
- number of active indices in a quadruply excited cluster
ACC options to scale various diagrams. All values are space separated.
- t2t2_t2:
- T_2^2 projected on doubles. Takes 5 values.
- t3_t2:
- T_3 projected on doubles. Takes 2 values.
- t1t3_t2:
- T_1 * T_3 projected on doubles. Takes 4 values.
- t2t2_t3:
- T_2^2 projected on triples. Takes 3 values.
- t2t3_t3:
- T_2 * T_3 projected on triples. Takes 5 values.