Nwchem 6.1 TCE error "tce hash: key not found 0"


Clicked A Few Times
Dear all,
I am trying to compute transition dipoles of a large system using keywords:


scf
thresh 1.0e-10
tol2e 1.0e-10
singlet
rhf
end
tce
freeze atomic
creom(t)ac
emin_act -1.0
emax_act 1.5
tilesize 15
thresh 1.0d-5
2eorb
2emet 6
idiskx 1
dipole
nroots 5
end
task tce energy

However I end up with lots of repetition of lines

after ga_get g_aom1
before ga_get g_aom1

Then lots of unexplained numbers, and finally the mentioned error
tce_hash: key not found 0

I googled, and found out that dipole moments were not implemented in version 5
Is it still the case for 6.1? The manual doesn't mention any limitation.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Dr. O. Baris Malcioglu,
University of Liege,
Bât. B5 Physique de la matière condensée
allée du 6 Août 17
4000 Liège 1
Belgique

Clicked A Few Times
Hi Baris,
The active-space CR-EOMCCSD(T) method calculates only the corrections to the active-space EOMCCSD excitation energies. If you want to calculate dipole moments you need to use either EOMCCSD or EOMCCSDT (we don't have dipole moments for the active-space EOMCCSD approach). Due to the size of the system, I guess EOMCCSD should be your method of choice. For the EOMCCSD calculations please use the following TCE settings

2eorb
2emet 14
split 2
tilesize 30
attilesize 40

These settings provide the best performance for the GA-based algorithm (using disk option may slow down the calculations). Anyway, 2eorb option is valid for the RHF and ROHF references (your case).
2emet 14 allows you to reduce the memory requirements during the 4-index transfromation by storing only a part of atomic 2-electron integrals. The "split" tells you that the original atomic 2-electron file will be split into 2 batches (only one batch is stored in memory at the time). You may used "split 3", "split 4" etc.

Best,
Karol


Forum >> NWChem's corner >> Running NWChem