Xi'an-CI Program
Contents
-
Xi'an-CI Program
- Corresponding email
- References
-
General keywords
- Electron
- nroot
- roots
- istate
- Symm
- Spin
- core
- Dele
- ORBTXT
- XvrUse
- QNEX
- Inactive
- Close
- Active
- SubDRT
- XSDSCI
- VSD
- NoVDVP
- MRCI
- Qss
- Qms
- H0Fock
- H0Dyall
- SAFOCK
- SSFOCK
- SDFOCK
- CSFiCI
- CFGiCI
- Cmin
- Pmin
- Qmin
- QminDV
- QminVD
- QminDD
- QminPV
- QminVP
- Qfix
- ICreduce
- NoICredu
- ReadDRT
- Nexci
- CVS
- GAS
- ReadREF
- SeleREF
- RootPrt
- Maxiter
- PRTCRI
- CITHR
- RShift
- VDRLS
- DVRLS
- DDRLS
- IShift
- VDILS
- DVILS
- DDILS
- DCRI
- EPIC
- EPCC
- Conv
- ETHRES
- InitHDav
- InitH0Dav
- FollowDav
- Memory keywords
- IC module keywords
- MRPT keywords
- Examples
Xi’an-CI program generates Multi Reference SDCI (MRCISD) wavefunctions (including internal contracted MRCISD on several different level accuracy), N-electron Valence states Second Order Perturbation Theory (including multi-state NEVPT2 (MS-NEVPT2), N-electron Valence states Third Order Perturbation Theory (NEVPT3), Static-Dynamic-Static Second Order Perturbation Theory (SDSPT2), Static-Dynamic-Static Configuration Interaction (SDSCI), Configuration Based Multi Reference Second Order Perturbation Theory (CBMRPT2) and Configuration Based Multi Reference Third Order Perturbation Theory (CBMRPT3). This program is based on hole-particle symmetry and GUGA for the computation of CI matrix elements. The program can calculate several eigenvectors simultaneously. Xi’an-CI program is written by Zhenyi Wen, Yubin Wang, Zhengting Gan, Bingbing Suo and Yibo Lei (Institute of Modern Physics, Northwest University, China).
Corresponding email
bsuo@nwu.edu.cn (Prof. Bingbing Suo) leiyb@nwu.edu.cn (Prof. Yibo Lei) wzy@nwu.edu.cn (Prof. Zhenyi Wen) yubin_wang@hotmail.com (Prof. Yubin Wang)
References
Xi’an-CI Program Package Review 1. B. Suo, Y. Lei, H. Han, Y. Wang, Mol. Phys., 116, 1051 (2018). ucMRCISD program 1. Y. Wang, G. Zhai, B. Suo, Z. Gan, Z. Wen, Chem. Phys. Lett., 375, 134 (2003). 2. Y. Wang, Z. Wen, Z. Zhang, Q. Du, J. Comput. Chem, 13, 187 (1992). 3. Y. Lei, B. Suo, Y. Dou, Y. Wang, Z. Wen, J. Comput. Chem, 31, 1752 (2010). 4. B. Suo, G. Zhai, Y. Wang, Z. Wen, X. Hu, L. Li, J. Comput. Chem, 26, 88 (2005). 5. Z. Gan, K. Su, Y. Wang, Z. Wen, Sci. China Ser. B-Chem, 42, 43 (1999). 6. Z. Wen, Y. Wang, H. Lin, Chem. Phys. Lett., 230, 41 (1994). icMRCISD program 1. Y. Wang, H. Han, Y. Lei, B. Suo, H. Zhu, Q. Song, Z. Wen, J. Chem. Phys., 141, 164114 (2014). 2. B. Suo, Y. Lei, H. Han, Y. Wang, Mol. Phys., 116, 1051 (2018). MS-NEVPT2 program 1. C. Angeli, R. Cimiraglia, S. Evangelisti, T. Leininger, J.P.Malrieu, J. Chem. Phys., 114, 10252 (2001). 2. Y. Lei, W. Liu, M. R. Hoffmann, Mol. Phys., 115, 2696 (2017). 3. B. Suo, Y. Lei, H. Han, Y. Wang, Mol. Phys., 116, 1051 (2018). SDSPT2 program 1. Y. Lei, W. Liu, M. R. Hoffmann, Mol. Phys., 115, 2696 (2017). 2. W. Liu, M.R. Hoffmann, Theor. Chem. Acc., 133, 1481 (2014). 3. W. Liu, M.R. Hoffmann, J. Chem. Theory Comput., 12, 1169 (2016); 12, 3000(E) (2016). CBMRPT2 program 1. Y. Lei, Y. Wang, H. Han, Q. Song, B. Suo, Z. Wen, J. Chem. Phys., 137, 144102 (2012). 2. A. Li, H. Han, B. Suo, Y. Wang, Z. Wen, Sci. China CHEMISTRY, 53. 933 (2010). 3. Y. Wang, Z. Gan, K. Suo, Z, Wen, Sci. China Ser. B-Chem, 43, 567 (2000).
General keywords
Comment:
If no keyword is used, xianci module will read information from mcscf and traint modules and then calculate Fully internal contracted MRCISD.
Electron
- CI effective electron Number without electrons of frozen MOs in traint module for MO integral transformation
Example:
Electron 30
nroot
- State Number, CASSCF with MixCI method needs to input state number of target CI type, which is equal to 'roots'
Example:
roots
- State Number, CASSCF with MixCI method needs to input state number of target CI type, which is equal to 'nroot'
Example:
istate
- Set the selected root index with total root number set firstly. The keyword supports only NEVPT2, SDSPT2, SDSCI and eSDSCI.
Example:
istate 2 !the total number of the selected roots 1 3 ! select the first and third roots.
Symm
- Symmetry of the target state, CASSCF with MixCI method needs to input irrep of target CI type.
Example:
Spin
- Spin multiplicity (2S+1), CASSCF with MixCI method needs to input Spin multiplicity of target CI type.
core
- Number of frozen or doubly occupied orbitals in each irreps. Default : no frozen orbitals.
Example:
Dele
- Number of deleted virtual orbitals in each irreps. Default : no deleted orbitals.
Example:
ORBTXT
set to read BDF_WorkDir text orbital file, such as $BDF_WORKDIR/$BDFTASK.inporb
Example:
Orbtxt inporb
XvrUse
The keyword for alternatively deleting virtual MOs by MCSCF XVR when keyword 'Dele' are not used to set delete MOs. Default is .false. However, the keyword 'Dele' is prior to 'XvrUse'.
QNEX
We defaultly approximately construct \bar{V}D and \bar{D}V subspaces by N{ex} {\bar{V}D}=1 and N{ex}_{\bar{D}V}=1. As this keyword is used N{ex}_{\bar{V}D}=2 and N{ex}_{\bar{D}V}=2.
Inactive
- Number of inactive orbitals in each irreps, which is equal to 'Close'
Example:
Close
- Number of inactive orbitals in each irreps, which is equal to 'Inactive'.
Example:
Active
- Number of active orbitals in each irreps.
Example:
Comment:
If the above keywords are not set. the mcscf and traint modules information will be used.
SubDRT
- active the function to parallel form sub-DRT. Default .TRUE.
Example:
Comment:
If the above keywords are not set. the mcscf and traint modules information will be used.
XSDSCI
- Calculate MRPT2 and MRCISD by SS-NEVPT2, MS-NEVPT2, SDSPT2+Q, SDSCI+Q and icMRCISD+Q. For icMRCISD, use MCSCF reference function and first-order wave function from purbutation theory as initial trial vectors to accelerate Davidson diagnoalization, deault : .false. This
keyword has the same funtion as the old keyword of eSDSCI.
Example:
XSDSCI
VSD
Use Virtual Space Decompostion (VSD) formed MOs to calculate XSDSCI, so that this keyword is combined with XSDSCI. VSD is used to separate Virtual MOs of large basis set by the selection of N(L)-N_occ of nonzero SVD value of <Psi(vir)|Phi(projected small basis set)>. This divide Virtual space into strong correlated space and weak correlated space, respectively.
Example: test126.inp
Notice : The number of virtual MOs by large basis set must be larger than the number of small basis set.
NoVDVP
Skip calculation of \bar(V)D and \bar(V)P for MRPT2.
Example: test148.inp
Notice : The number of virtual MOs by large basis set must be larger than the number of small basis set.
MRCI
- Calculate MRCISD where the reference CSFs are default optimized with the assistance of internal contraction mode. Deault : .true.
Example:
MRCI
Qss
- Use state-specific Q space for NEVPT2, SDSPT2 and SDSCI, default : .true.
Qms
- Use state-universal Q space for NEVPT2, SDSPT2 and SDSCI, default : .false.
H0Fock
Use generalized Fock operator as zeroth-order Hamiltonian (H0) to calculate <Q|H0|Q>, which is remarkably quicker than that use Dyall Hamiltonian as H0. Default : .false.
H0Dyall
Use Dyall Hamiltonian as zeroth-order Hamiltonian (H0) to calculate <Q|H0|Q>, which is remarkably slower than that use generalized Fock operator as H0. Default : .true.
SAFOCK
- Use state-average CMO energies and integrals for NEVPT2, SDSPT2 and SDSCI, default : .true.
SSFOCK
- Use state-dependent CMO energies and integrals for NEVPT2, SDSPT2 and SDSCI, default : .false.
SDFOCK
- Use state-dependent CMO energies and state-average CMO integrals for NEVPT2, SDSPT2 and SDSCI, default : .false.
CSFiCI
- This keyword can set to use reference CSF but not oCFG. If keyword 'CFGiCI' on MCSCF or Xianci module is set, this functional does not work. When 'CSFiCI' on Xianci module is set, it works. When it is active, users can use iCISCF with respect to reference CSF.
CFGiCI
- This keyword can set to use reference oCFG but not CSF. Default : .false. If keyword 'CFGiCI' on MCSCF or Xianci module is set, this functional works. When it is active, users can use iCISCF with respect to reference oCFG which can be used by keyword CFGiCI on MCSCF.
Cmin
- Threshold of CI coefficient for reference CSFs when use keyword 'CSFiCI', the Cmin value can be read from MCSCF for iCISCF(2), or users can set this keyword manually. Cmin can be used to select reference CSF but not oCFG after the pre-calculation of H0 diagonalization with respect to reference oCFG.
Pmin
- Threshold of CI coefficient for reference CSFs which will read from $WORKDIR/$BDFTASK.select_*_#, where * is the spin multiplicity, # is the irreducible representation. Default : 0.0. The input threshold should be larger than the default 0.0. When use this keyword the P space will be enlarged and E(H0) will use the enlarged P space energy and iCIPT2 energy will be set to zero.
When do not use this keyword the E(H0) energy is equal to iCISCF energy and iCIPT2 energy will be added to the following MRCISD or MRPT2. This keyword is equal to 'CSFCRI'.
Qmin
Threshold of uniquely pruning Q subspaces to reduce internally contracted wavefunctions (it|uv),(tu|va),(iu|va), (ij|uv) and (uv|ab) by semistochastic heat-bath CI (SBCI) with max<q|H|0> < Qmin, default : 0.d0.
QminDV
Threshold of internal contraction coefficient to reduce \bar(D)V internally contracted wavefunctions (it|uv) by semistochastic heat-bath CI (SBCI) with max<q|H|0> < QminDV, default : 0.d0.
QminVD
Threshold of internal contraction coefficient to reduce \bar(V)D internally contracted wavefunctions (tu|va) by semistochastic heat-bath CI (SBCI) with max<q|H|0> < QminVD, default : 0.d0.
QminDD
Threshold of internal contraction coefficient to reduce \bar(D)D internally contracted wavefunctions (iu|va) by semistochastic heat-bath CI (SBCI) with max<q|H|0> < QminDD, default : 0.d0.
QminPV
Threshold of internal contraction coefficient to reduce \bar(P)V internally contracted wavefunctions (ij|uv) by semistochastic heat-bath CI (SBCI) with max<q|H|0> < QminPV, default : 0.d0.
QminVP
Threshold of internal contraction coefficient to reduce \bar(V)P internally contracted wavefunctions (uv|ab) by semistochastic heat-bath CI (SBCI) with max<q|H|0> < QminVP, default : 0.d0.
Qfix
- Threshold to Fix CI coefficients of Q space for ic-MRCISD diagonalization, default : 0.d0.
ICreduce
- This keyword is used to reduce \bar(D)V and \bar(V)D internally contracted wavefunctions by merging direct and exchanged ploops, so that the pair two IC functions with the small active orbital indexes are merged together. Default : .true.
NoICredu
- This keyword is used to do not reduce \bar(D)V and \bar(V)D internally contracted wavefunctions by merging direct and exchanged ploops, so that the pair two IC functions with the small active orbital indexes are merged together. Default : .false.
ReadDRT
- read DRT from $WORKDIR/$BDFTASK.cidrt Default is .false.
Nexci
- set excitation number relative to reference CSFs, default = 2 for MRCISD/MRPT2. If Nexci=1 for MRCIS or MRPT2 with only one electron excitation.
Example:
Nexci 1
CVS
Core Valence Separation for Core excitation for GUGA if use this keyword. Default = .false.
Example:
GAS
several lines should be provided for controlling GASSCF calculations. Default is read from MCSCF and needs not to set. Line 1: number for GAS spaces, like GAS1, GAS2, GAS3, .... Line 2: minimum electron occupation numbers for the GAS spaces. Line 3: maximum electron occupation numbers for the GAS spaces. From Line 4 to Line (GAS spaces number plus 3) set active orbital with symmetry of these GAS spaces.
Example:
gas 2 ! there are two GAS spaces. 2 4 ! minimum electron occupation numbers for the GAS spaces. 4 10 ! maximum electron occupation numbers for the GAS spaces. 2 0 0 0 ! active orbitals of each irreps of GAS1 2 0 2 2 ! active orbitals of each irreps of GAS2.
Comment:
With keyword 'GAS' setting, keywords 'active' is useless and can be missing.
ReadREF
- Automatically read REF CSF from $WORKDIR/$BDFTASK.select_*_#, where * is the spin multiplicity, # is the irreducible representation. The functional is combined with keyword 'CSFCRI' to read and select reference CSF from text file.
Example:
$xianci ... READREF ... $end
SeleREF
- Line 1: Set Number of Selected CSF occupations (Nref). Line 2 to Line Nref+1: set occupation (2,1,0) respectively to double, single and zero occupation.
Example:
SELEREF 3 2200 2110 2020
RootPrt
- Print the target state (root) energy for calculating numerical gradient of this state in numgrad module, default is 1.
Example:
RootPrt 3 # the third state (root) energies will be printed.
Maxiter
- Maximum iteration Number of Davidson Diagonalization used in Xi'an-CI. The default value is 200.
Example:
Maxiter 50
PRTCRI
- set threshold for CI vector print, which is equal to keyword 'CITHR'. The default value is 0.05.
CITHR
- set threshold for CI vector print, which is equal to keyword 'PRTCRI'. The default value is 0.05.
Example:
CITHR 0.1
RShift
set real level shift for <Q|H(0)-E(0)+Rshift|Q>. The default value is 0.d0. Recommend: 0.3
Example:
Rshift 0.3
VDRLS
set real level shift for <\bar{V}D|H(0)-E(0)+VDRLS |\bar{V}D>. The default value is 0.d0. Recommend: 0.3
Example:
VDRLS 0.3
DVRLS
set real level shift for <\bar{D}V|H(0)-E(0)+DVRLS |\bar{D}V>. The default value is 0.d0. Recommend: 0.3
Example:
DVRLS 0.3
DDRLS
set real level shift for <\bar{D}D|H(0)-E(0)+DDRLS |\bar{D}D>. The default value is 0.d0. Recommend: 0.3
Example:
DDRLS 0.3
IShift
set imaginary level shift for <Q|H(0)-E(0)+Ishift|Q>. The default value is 0.d0. Recommend: 0.3
Example:
Ishift 0.3
VDILS
set imaginary level shift for <\bar{V}D|H(0)-E(0)+VDILS |\bar{V}D>. The default value is 0.d0. Recommend: 0.3
Example:
VDILS 0.3
DVILS
set imaginary level shift for <\bar{D}V|H(0)-E(0)+DVILS |\bar{D}V>. The default value is 0.d0. Recommend: 0.3
Example:
DVILS 0.3
DDILS
set imaginary level shift for <\bar{D}D|H(0)-E(0)+DDILS |\bar{D}D>. The default value is 0.d0. Recommend: 0.3
Example:
DDILS 0.3
DCRI
- set threshold for internal contracted CSF (ICCSF) deleting and linear-dependent orthonormalization. The default value is 1.d-12.
Example:
CITHR 1.d-12
EPIC
- set threshold for internal contracted coefficient which are used to form Hamiltonian matrices. The default value is 0.d0. Recommend: 1.d-5
Example:
epic 1.d-5
EPCC
- set threshold for CI coupling coefficient which are used to form Hamiltonian matrices. The default value is 0.d0. Recommend: 1.d-10
Example:
epcc 1.d-10
Conv
- set threshold for CI energy, CI vector and Residual vector of MRCISD, respectively. The default value is set as the following example.
Example:
Conv 1.d-8 1.d-4 1.d-8
ETHRES
- set threshold for CI energy of H0. The default value is 1.d-8 .
InitHDav
- Set 1 for initial vectors on Davidson diagonalization by largest coupling with energy lowest CSFs,which is default. Set 2 for initial vectors on Davidson diagonalization by low-lying CI Hamiltonian diagonal elements near by reference states. Set 3 for initial vectors on Davidson diagonalization by the residual vector of reference wavefunctions.
InitH0Dav
- Set 1 for initial vectors of P space on Davidson diagonalization by largest coupling with energy lowest CSFs for the H0 diagonalization. Set 2 for initial vectors of P space on Davidson diagonalization by low-lying CI Hamiltonian diagonal elements near by reference CSFs for the H0 diagonalization, which is default.
FollowDav
- If this keyword is specified in a MRCISD calculation, and if the $numgrad block is present, the numerical gradient calculation will use the MRCISD+Q energies. Otherwise the uncorrected MRCISD energies will be used.
Memory keywords
Nosavelp
- set to calculate partial loops on each MRCI iteration individually. This leads to more MRCI calculation time but saves hard disk and should be used to the system with large active space.
H0Tra
- Maximum H0 dimension can completely diagonalize H0 matrix. The default value is 1000.
NCISAVE
- Maximum H0 dimension can save H0 matrix, which insteads of the old keyword of 'H0TRA'. The default value is 50000.
MAXREF
- Maximum selected reference CSF number. The default value is 50000.
Example:
NODE 50000
NODE
- Maximum DRT node number. The default value is 1000000.
Example:
NODE 1000000
subNODE
- Maximum sub-DRT node number. The default value is 1000000.
Example:
NODE 100000
Maxload
- Set largest number of IC coefficient of each block. Default: 10000000000.
PLBLK
- Maximum partial LOOP block number. The default value is 2000000.
Example:
PLBLK 10000000
IC module keywords
UCCI
- This keyword is set for un-contracted MRCISD.
Example:
UCCI
FCCI
- Default for internal contraction module. This keyword is set for Fully internal Contraction module of CSFs, reference CSFs are not contracted for MRCISD calculation, while perturbation theory calculation all CI subspaces are internally contracted.
Example:
FCCI
NICI
- This keyword is set for one internal Contraction module of CSFs, only internal CI subspaces are not contracted.
Example:
NICI
CWCI
- This keyword is set for one internal Contraction module of CSFs, corresponding to keyword 'mrcic' in Molpro program for Celani-Werner (CW) contraction, where only CI subspaces VV, DV, DDV and VD in hole-particle symmetry are not contracted.
Example:
CWCI
WKCI
- This keyword is set for one internal Contraction module of CSFs, corresponding to keyword 'mrci' in Molpro program for Werner-Knowles (WK) contraction, where only CI subspaces with two electron excitation to external spaces are contracted.
Example:
WKCI
SDCI
- This keyword is set for one internal Contraction module of CSFs, the accuracy of which is more accurate than CWCI but less than WKCI. In contrast with WKCI module, CI subspaces with two electron excitation from hole space and meanwhile one electron excitation to external space are also contracted.
Example:
SDCI
XDCI
- This keyword is set for one internal Contraction module of CSFs, where XS(T), S(T)V and S(T)D are not contracted.
Example:
XDCI
MRPT keywords
Comment:
If no keyword is set for perturbation theory calculation in the following, xianci module will calculate MRCISD in default.
Notice:
The following methods use Fully internal contraction wavefunction (FCCI) as default, while NICI, CWCI, SDCI, WKCI modules should be set in turn for the case FCCI module fails.
NEVPT2
- set for SS-NEVPT2 and MS-NEVPT2 calculations, where each reference state expands a specific CI space.
Example:
NEVPT2
MR-NEVPT2
- set for SS-NEVPT2 and MS-NEVPT2 calculations, where all reference states expand only one multi-states CI space.
Example:
MR-NEVPT2
NEVPT3
- set for SS-NEVPT3 calculation, where each reference state expands a specific CI space.
Example:
NEVPT3
SDSPT2
- set for SDSPT2 calculation, where all reference states expand only one multi-states CI space.
Example:
SDSPT2
SDSCI
- set for SDSCI calculation, where all reference states expand only one multi-states CI space.
Example:
SDSCI
DYLAN
set for SDSPT2 and SDSCI calculations, where truncated psi(0)_i dynamically combinated Lanczos Psi2 by sum_i<psi(0)_i|H|Psi(1)> are used to generate Ps wavefunction in SDSPT2 and SDSCI. Default = .true.
NOLAN
- set for SDSPT2 and SDSCI calculations, where high-lying MCSCF wavefunction as Psi2 are used to generate Ps wavefunction in SDSPT2 and SDSCI. Default = .false.
DOLAN
- set for SDSPT2 and SDSCI calculations, where Lanczos wavefunction as Psi2 are used to generate Ps wavefunction in SDSPT2 and SDSCI. Default = .false.
DEPSI2
- set threshold for the cutoff of the calculated number of H0 states, those of which are high-lying than the target states. The default value is 0.5 eV, users can set the threshold with unit of eV.
NDIMPS
- set for SDSPT2 and SDSCI calculations, where CASSCF wavefunctions are used to produce Ps wavefunction in SDSPT2 and SDSCI.
Example:
NDIMPS 2 # two high-lying CASSCF wavefunctions are used to produce Ps wavefunction in SDSPT2 and SDSCI relative to reference wavefunctions.
Comment:
If Keyword 'NDIMPS' are not set or set to zero and keyword 'NOLAN' are set, SDSPT2 or SDSCI has no Ps wavefunction.
CBMRPT2
- set for CB-MRPT2 calculation, where each reference state expands a specific CI space.
Example:
CBMRPT2
MR-CBMRPT2
- set for CB-MRPT2 calculations, where all reference states expand only one multi-states CI space.
Example:
MR-CBMRPT2
MR-CBMRPT3
- set for CB-MRPT3 calculations, where all reference states expand only one multi-states CI space.
Example:
MR-CBMRPT3
Examples
Test Example 1
input:
$COMPASS Title C2H4 Molecule test run Basis cc-pvdz Geometry C 0.000000 1.386400 0.000000 C 0.000000 -1.386400 0.000000 C 2.099700 2.794200 0.000000 C -2.099700 -2.794200 0.000000 H -1.845200 2.307000 0.000000 H 1.845200 -2.307000 0.000000 H 3.968500 1.930200 0.000000 H -3.968500 -1.930200 0.000000 H 2.015100 4.847500 0.000000 H -2.015100 -4.847500 0.000000 END geometry Check unit bohr $END $xuanyuan $end $SCF RHF charge 0 spin 1 $END $MCSCF close 7 0 0 5 active 0 2 3 1 actele 6 spin 1 symmetry 1 roots 3 3 1 2 3 1 1 1 mixci 2 1 3 2 1 1 4 ROOTPRT 1 prtcri 0.1 guess hforb $END $TRAINT Frozen 2 0 0 2 0 0 0 0 Orbital mcorb $END $XIANCI nroot 2 spin 1 symmetry 1 $END $XIANCI nroot 1 spin 3 symmetry 4 $END
Results:
========================= mcscf results ============================== State Averaged ci energy -154.86258790 root 1 energy= -154.98691206 exe(eV)= 0.0000 root 2 energy= -154.73707954 exe(eV)= 6.7983 root 3 energy= -154.86377210 exe(eV)= 3.3508 ++++++++ DATA CHECK +++++++++++++++++++++++++++++++++ CHECKDATA:MCSCF:MCENERGY: -154.9869121 -154.7370795 -154.8637721 ++++++++++ END DATA CHECK ++++++++++++++++++++++++++++ End MCSCF Calculation ========================= xianci results ============================== =============================== For first type of CI with two singlet states ==================================== Roots of Heff are calculated are listed below: ENE ENE + Pople ENE + App Pople ENE + DAV ENE + MEISS root 1 -155.45209027 -155.52854668 -155.52960628 -155.51383149 -155.51395190 root 2 -155.19957647 -155.27731997 -155.27842584 -155.26200965 -155.26229526 MRCISD energyies Pople Correction App Pople Correction Davidson Correction Meissner correction ===================================================== MRSDCI CALCULATION CONVERGED NROOT MC ENERGY CI ENERGY CI DAV DAVCOEF 1 -154.98691206 -155.45209027 -155.51383149 0.867274 2 -154.73707954 -155.19957647 -155.26200965 0.865008 MCSCF energyies MRCISD energyies Davidson Correction Reference weight root 1 energy= -155.45209027 exe(eV)= 0.0000 root 2 energy= -155.19957647 exe(eV)= 6.8713 ++++++++ DATA CHECK +++++++++++++++++++++++++++++++++ CHECKDATA:MRCI:CIENERGY: -155.4520903 -155.1995765 ++++++++++ END DATA CHECK ++++++++++++++++++++++++++++ =============================== For second type of CI with one triplet state ==================================== Roots of Heff are calculated are listed below: ENE ENE + Pople ENE + App Pople ENE + DAV ENE + MEISS root 1 -155.32503309 -155.40089070 -155.40194273 -155.38628185 -155.38640551 ===================================================== MRSDCI CALCULATION CONVERGED NROOT MC ENERGY CI ENERGY CI DAV DAVCOEF 1 -154.86377210 -155.32503309 -155.38628185 0.867215 root 1 energy= -155.32503309 exe(eV)= 0.0000 ++++++++ DATA CHECK +++++++++++++++++++++++++++++++++ CHECKDATA:MRCI:CIENERGY: -155.3250331 ++++++++++ END DATA CHECK ++++++++++++++++++++++++++++
Test Example 2
input:
$TRAINT Frozen 2 0 0 2 0 0 0 0 Orbital mcorb mrpt2 $END $XIANCI nroot 2 spin 1 symmetry 1 SDSPT2 $END
Results:
=============================== For first type of CI with two singlet states ==================================== NROOT MC ENE SS-NEVPT2 ENE MS-NEVPT2 ENE SDSPT2 ENE SDSPT2+Q1 ENE SDSPT2+Q2 ENE SDSPT2+Q3 ENE DAVCOEF 1 -154.98691206 -155.47745410 -155.47745446 -155.41455599 -155.47503759 -155.47574313 -155.46512580 0.881748 2 -154.73707954 -155.21961390 -155.21961354 -155.15793413 -155.21775988 -155.21846183 -155.20789974 0.881276 Energies: MCSCF SS-NEVPT2 MS-NEVPT2 SDSPT2 Pople Correction App Pople Correction Davidson Correction Ref. Weight
Test Example 3
input:
$XIANCI nroot 2 spin 1 symmetry 1 SDSCI $END
Results:
=============================== For first type of CI with two singlet states ==================================== NROOT MC ENE SS-NEVPT2 ENE MS-NEVPT2 ENE SDSPT2 ENE SDSPT2+Q1 ENE SDSPT2+Q2 ENE SDSPT2+Q3 ENE DAVCOEF 1 -154.98691206 -155.47745410 -155.47745446 -155.44006672 -155.51313986 -155.51413050 -155.49935009 0.869176 2 -154.73707954 -155.21961390 -155.21961354 -155.18843582 -155.26361048 -155.26466844 -155.24894428 0.865941 Energies: MCSCF SS-NEVPT2 MS-NEVPT2 SDSCI Pople Correction App Pople Correction Davidson Correction Ref. Weight
Test Example 4
input:
$XIANCI nroot 2 spin 1 symmetry 1 NEVPT3 $END
Results:
=============================== For first type of CI with two singlet states ==================================== NROOT MC ENERGY SS-NEVPT2 ENERGY MS-NEVPT2 ENERGY SS-NEVPT3 ENERGY MS-NEVPT3 ENERGY 1 -154.98691206 -155.47742562 -155.47742574 -155.51364676 -155.51364676 2 -154.73707954 -155.21952164 -155.21952152 -155.26247430 -155.26247430 Energies: MCSCF SS-NEVPT2 MS-NEVPT2 SS-NEVPT3 Useless
Test Example 5
input:
$XIANCI nroot 2 spin 1 symmetry 1 CBMRPT2 $END
Results:
=============================== For first type of CI with two singlet states ==================================== ++++++++ DATA CHECK +++++++++++++++++++++++++++++++++ CHECKDATA:MRPT2:PT2ENERGY: -155.5496768 -155.2931467 ++++++++++ END DATA CHECK ++++++++++++++++++++++++++++
Test Example 6
input:
$XIANCI nroot 2 spin 1 symmetry 1 MR-CBMRPT3 $END
Results:
=============================== For first type of CI with two singlet states ==================================== ++++++++ DATA CHECK +++++++++++++++++++++++++++++++++ CHECKDATA:MRPT3:PT3ENERGY: -155.5176000 -155.2629435 ++++++++++ END DATA CHECK ++++++++++++++++++++++++++++