Basis set error, input is inconsistent!

User12 · 2020-10-22 11:56:09

Dear All,
I am doing some initio calculations of certain Lanthanide complexes, while doing the GuesOrb calculations i am facing some problems with the basis sets.

MPI_ABORT was invoked on rank 13 in communicator MPI_COMM_WORLD
with errorcode 112.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    ERROR:  input is inconsistent!                                       ###
 ###    SEWARD found basis sets of mixed relativistic (or non-relativistic)  ###
 ###    types!                                                               ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################
--- Stop Module: gateway at Tue Oct 20 12:44:47 2020 /rc=_RC_INPUT_ERROR_ ---
*** files: xmldump
    saved to directory /leviathan/home/afpcomba06/dy
--- Module gateway spent 6 seconds ---

.####################.
.# User input error #.
.####################.

    Timing: Wall=6.08 User=10.65 System=36.97

I think the problem is, that I want to use more than one basis set. I also read something about the XBAS Keyword, does someone have experience with it?

My input

> export MOLCAS_MOLDEN=ON
 &GATEWAY
Title
Dy_normal
Charge
1
Coord
Dy-opt.xyz
Basis set
Dy.ANO-RCC...8s7p5d3f2g1h.,N.ANO-RCC...4s3p2d1f.,O.ANO-RCC...4s3p2d1f.,C.ANO-RCC...3s2p.,H.ANO-RCC...2s.

ANGM
8.800080097      3.501473927     14.749932517
SDIPolar
AMFI
Douglas-Kroll

&SEWARD
Cholesky
&GUESSORB
PRMO
3
PRPOpulation

End of input

Thanks in advance for your help.
Alina

nikolay · 2020-10-22 13:02:53

Hi Alina,

I remember having the same trouble.
I guess that the long string after Basis set got silently truncated and the default ANO-L or ANO-S is used for some of the elements at the end.
You should see that scrutinizing the output of gateway.
There should be some way to give longer basis set or other lines, but I don't know it.
Ignacio, do you have an idea?

Ignacio · 2020-10-22 14:58:50

That could be the problem, but the line doesn't look long enough for truncation (I think the limit is 180 characters). Maybe you have other elements in the xyz file that have no specified basis and get the default.

You can specify a default basis, and specific ones, and you can use shorthand forms (to avoid truncation), and you can write the basis directly in the xyz file for each atom. For instance:

Basis = ANO-RCC-VTZP,C.ANO-RCC-VDZ,H.ANO-RCC-VDZ

User12 · 2020-10-23 10:59:01

Thanks for the responses. I tryd it again with the shorthand forms of the basis sets, but there is again an Error. I'm sure, that the corresponding basis sets are installed.

MPI_ABORT was invoked on rank 10 in communicator MPI_COMM_WORLD
with errorcode 112.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    ERROR: Basis set file /work/install/local/basis_library/ANO-RCC-TZP  ###
 ###    does not exist!                                                      ###
 ###                                                                         ###
 ###    (1) For a valence basis set: check the spelling of                   ###
 ###    the basis set label and that the basis set file is                   ###
 ###    present in the basis set library directory.                          ###
 ###                                                                         ###
 ###    (2) For an external auxiliary basis set: check that                  ###
 ###    the basis set file is present in the appropiate basis                ###
 ###    set library subdirectory.                                            ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################
--- Stop Module: gateway at Fri Oct 23 11:47:07 2020 /rc=_RC_INPUT_ERROR_ ---
*** files: xmldump
    saved to directory /leviathan/home/afpcomba06/test2_coord
--- Module gateway spent 5 seconds ---
                                                                                                                  131,2         95%

My input

> export MOLCAS_MOLDEN=ON
 &GATEWAY
Title
test_D
Charge
1
Coord
test_D.xyz

Basis = Dy.ANO-RCC-TZP,N.ANO-RCC-TZP,O.ANO-RCC-TZP,C.ANO-RCC-DZP,H.ANO-RCC-VDZ


ANGM
8.800080097      3.501473927     14.749932517
SDIPolar
AMFI
Douglas-Kroll

&SEWARD
Cholesky
&GUESSORB
PRMO
3
PRPOpulation

End of input

Thanks
Alina

nikolay · 2020-10-23 13:32:07

I think DZP is not a correct abbreviation, you can always verify with

pymolcas help_basis Dy

Following the short notation of Ignacio, It should read

Basis = ANO-RCC-VTZP,C.ANO-RCC-VDZP,H.ANO-RCC-VDZ

Previously you had C.ANO-RCC...3s2p, which is VDZ not VDZP

pymolcas help_basis C
...
C.ANO-RCC-VDZ -> C.ANO-RCC...3s2p
C.ANO-RCC-VDZP -> C.ANO-RCC...3s2p1d
...

User12 · 2020-10-27 11:28:41

Thanks for your help.
I use openmolcas and the command doesen't work for me. Do you know what I can use for openmolcas to verify the abbreviation?

Thanks in advance for your help.

Ignacio · 2020-10-27 14:30:36

What nikolay said works in OpenMolcas. If it doesn't work for you, you may have a broken or misconfigured installation. In any case, you can manually check the file $MOLCAS/basis_library/basis.tbl

User12 · 2020-11-10 23:08:49

Thanks again for your help.
I tried two different inputs for the guessOrb calculation, which should lead to the same result. The guessOrb calculation took about the same time for both inputs and the output looks similar.
My input for the guessOrb (long)

> export MOLCAS_MOLDEN=ON
 &GATEWAY
Title
Dy_long
Charge
1
Basis set
Dy.ANO-RCC...8s7p5d3f2g1h.
Dy1       xyz
End of basis

Basis set
N.ANO-RCC...4s3p2d1f.
xyz of N
End of basis

Basis set
O.ANO-RCC...4s3p2d1f.
xyz of O

End of basis

Basis set
C.ANO-RCC...3s2p.
xyz of C

End of basis

Basis set
H.ANO-RCC...2s.
xyz of H


End of basis

ANGM
xyz
SDIPolar
AMFI
Douglas-Kroll

&SEWARD
Cholesky
&GUESSORB
PRMO
3
PRPOpulation

End of input

My input for the guessOrb (short)

> export MOLCAS_MOLDEN=ON
 &GATEWAY
Title
Dy_short
Charge
1
Coord
Dy_short.xyz
Basis = ANO-RCC-VTZP,C.ANO-RCC-VDZ,H.ANO-RCC-VDZ.

ANGM
xyz
SDIPolar
AMFI
Douglas-Kroll

&SEWARD
Cholesky
&GUESSORB
PRMO
3
PRPOpulation

End of input

After that I started the rasscf calculation and am wondering why the input with the short form takes significantly longer (2 days so far) than the other.
I see that the calculation is still at the Wave function control section and that it requires significantly more iterations to converge than with the other input.
I’m just wondering whether I did something wrong with the input for the rasscf.

My rasscf input

&RASSI
Nr of JobIph
1 21
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Spin
MEES
Properties
3
'AngMom' 1 'AngMom' 2 'AngMom' 3
SOProperties
3
'AngMom' 1 'AngMom' 2 'AngMom' 3
EJOB

Thanks in advance for your help.
Alina

Ignacio · 2020-11-11 09:32:09

That's RASSI, not RASSCF.

One difference is that the "long" version defaults to no symmetry, while the "short" version defaults to full symmetry. Then the atoms and basis functions could be in different order, which may be important for the RASSCF calculation, depending on how you specify the initial orbitals and active space.

User12 · 2020-11-11 09:46:11

Sorry my bad I uploaded the wrong imput.
Here is the one I intendet to show.

> export MOLCAS_MOLDEN=ON
&RASSCF
Spin
6
Inactive
225
Nactel
9 0 0
Ras2
7
Alter
7
1 226 211
1 227 212
1 228 213
1 229 214
1 230 215
1 231 217
1 232 218
CiRoot
21 21 1
> COPY $Project.JobIph JOB001

Thanks for the helpful comment. I suppose I have to modify my rasscf input for the "short" version, but I'm not sure in which way.
I'd appreciate some help in this regard.

Ignacio · 2020-11-11 09:54:19

Possibly the orbital numbers for the Alter keyword should be different. Check the guess orbitals generated with the "short" version. Otherwise, make sure you are not using the guess orbitals from one version with the other if the atoms are in different order.

User12 · 2020-11-11 11:01:26

Thanks again for your help.

My supervisor asked me to fix the "short" input to default to no symmetry.
I adjusted the input as follows:

> export MOLCAS_MOLDEN=ON
 &GATEWAY
Title
Dy1-re
Charge
1
Coord
Dy1-re.xyz
Group=nosym
Basis = ANO-RCC-VTZP,C.ANO-RCC-VDZ,H.ANO-RCC-VDZ

ANGM
8.76570000    3.40420000   14.76950000
SDIPolar
AMFI
Douglas-Kroll

&SEWARD
Cholesky
&GUESSORB
PRMO
3
PRPOpulation

End of input

Molcas Forum

Announcement

#1 2020-10-22 11:56:09

Basis set error, input is inconsistent!

#2 2020-10-22 13:02:53

Re: Basis set error, input is inconsistent!

#3 2020-10-22 14:58:50

Re: Basis set error, input is inconsistent!

#4 2020-10-23 10:59:01

Re: Basis set error, input is inconsistent!

#5 2020-10-23 13:32:07

Re: Basis set error, input is inconsistent!

#6 2020-10-27 11:28:41

Re: Basis set error, input is inconsistent!

#7 2020-10-27 14:30:36

Re: Basis set error, input is inconsistent!

#8 2020-11-10 23:08:49

Re: Basis set error, input is inconsistent!

#9 2020-11-11 09:32:09

Re: Basis set error, input is inconsistent!

#10 2020-11-11 09:46:11

Re: Basis set error, input is inconsistent!

#11 2020-11-11 09:54:19

Re: Basis set error, input is inconsistent!

#12 2020-11-11 11:01:26

Re: Basis set error, input is inconsistent!

Board footer