× If you have problems during the execution of MRCC, please attach the output with an adequate description of your case as well as the followings:
  • the way mrcc was invoked
  • the way build.mrcc was invoked
  • the output of build.mrcc
  • compiler version (for example: ifort -V, gfortran -v)
  • blas/lapack versions
  • as well as gcc and glibc versions

This information really helps us during troubleshooting :)

RHF Discrepancy between MRCC and MOLPRO

More
6 years 4 months ago #291 by Nike
Dear Colleagues,

I am getting very different RHF energies in MOLPRO vs stand-alone MRCC, when using the same basis set. It looks like MRCC is finding a local minimum, because the energy is higher than MOLPRO's, and usually this can be fixed by specifying the occupation. However, in both cases the occupation is (10,2):
# TITLE
basis=aug-cc-pVDZ-EMSL
calc=CISD
mem=8GB
core=corr
ccmaxit=999
scfmaxit=9999

geom
O
O 1 r1
O 2 r2 1 theta

r1=2.28188
r2=4.5
theta=116.8

unit=bohr
O:aug-cc-pVDZ-EMSL
EMSL BASIS SET LIBRARY

  3
    0    1    2
    4    3    2
   10    5    2

 11720.0000000  1759.0000000   400.8000000   113.7000000    37.0300000
    13.2700000     5.0250000     1.0130000     0.3023000     0.0789600

 0.0007100 -0.0001600  0.0000000  0.0000000
 0.0054700 -0.0012630  0.0000000  0.0000000
 0.0278370 -0.0062670  0.0000000  0.0000000
 0.1048000 -0.0257160  0.0000000  0.0000000
 0.2830620 -0.0709240  0.0000000  0.0000000
 0.4487190 -0.1654110  0.0000000  0.0000000
 0.2709520 -0.1169550  0.0000000  0.0000000
 0.0154580  0.5573680  0.0000000  0.0000000
-0.0025850  0.5727590  1.0000000  0.0000000
 0.0000000  0.0000000  0.0000000  1.0000000

    17.7000000     3.8540000     1.0460000     0.2753000     0.0685600

 0.0430180  0.0000000  0.0000000
 0.2289130  0.0000000  0.0000000
 0.5087280  0.0000000  0.0000000
 0.4605310  1.0000000  0.0000000
 0.0000000  0.0000000  1.0000000

     1.1850000     0.3320000

 1.0000000  0.0000000
 0.0000000  1.0000000

E_RHF,MRCC = -224.2206354441273390
This is the same if I add symm=1.


*** title
memory,8400,m
 geometry={
O1;
O2, O1, r1;
O3, O1, r2, O2, theta
}


r1=2.28188
r2=4.5
theta=116.8

basis={
!
! OXYGEN       (10s,5p,2d) -> [4s,3p,2d]
! OXYGEN       (9s,4p,1d) -> [3s,2p,1d]
! OXYGEN       (1s,1p,1d)
s, O , 11720.0000000, 1759.0000000, 400.8000000, 113.7000000, 37.0300000, 13.2700000, 5.0250000, 1.0130000, 0.3023000, 0.0789600
c, 1.8, 0.0007100, 0.0054700, 0.0278370, 0.1048000, 0.2830620, 0.4487190, 0.2709520, 0.0154580
c, 1.8, -0.0001600, -0.0012630, -0.0062670, -0.0257160, -0.0709240, -0.1654110, -0.1169550, 0.5573680
c, 9.9, 1
c, 10.10, 1
p, O , 17.7000000, 3.8540000, 1.0460000, 0.2753000, 0.0685600
c, 1.3, 0.0430180, 0.2289130, 0.5087280
c, 4.4, 1
c, 5.5, 1
d, O , 1.1850000, 0.3320000
c, 1.1, 1
c, 2.2, 1
}

print basis,orbitals

{rhf;wf,24,1,0}

rcisd

E_RHF,MOLPRO = -224.231538303312

Is there any other reason why MRCC might be giving me a local minimum?

Please Log in or Create an account to join the conversation.

More
6 years 4 months ago #292 by kallay
Dear Nike,
If the SCF does not find the global minimum, you should play a bit with the initial guess or the convergence acceleration procedures, because it may resolve the problem. In your case, if I change the initial guess to scfiguess=ao, I get the molpro energy.

Best regards,
Mihaly Kallay

Please Log in or Create an account to join the conversation.

More
6 years 4 months ago #293 by Nike
Great answer!
Thank you, I will find this useful for many systems in the future.
With best wishes,
Nike Dattani

Please Log in or Create an account to join the conversation.

Time to create page: 0.016 seconds
Powered by Kunena Forum