Page 1 of 1

Basis set label problem

Posted: 21 Nov 2013, 08:50
by Rusakur
At the geometry optimization the following error occurs: the name of basis set is transformed into the capital letters.

Input:
Content of the .dal input file
----------------------------------

**DALTON INPUT
.PRESORT
.OPTIMIZE
*MOLBAS
.SYMTHR
1D-15
**INTEGRALS
**WAVE FUNCTIONS
.DFT
PBE0
*END OF INPUT


Content of the .mol file
----------------------------

ATOMBASIS
Input draft: C1
C1 symmetry, basis cc-pVDZ
Angstrom Atomtypes=3 Charge=0
Charge=34 Atoms=1 Basis=cc-pVDZ
Se2 -0.316200 0.020371 -0.000469
Charge=6 Atoms=1 Basis=cc-pVDZ
C1 1.641687 -0.016260 0.000244
Charge=1 Atoms=4 Basis=cc-pVDZ
H3 2.015759 -0.486479 0.908439
H4 2.015587 -0.530473 -0.883859
H5 1.962901 1.025578 -0.025359
H6 -0.468156 -1.430362 0.035042

And output:
**************************************************************************************
****************** Output of molecule and basis set info suppressed ******************
**************************************************************************************


Basis "CC-PVDZ " does not exist.

The following directories have been searched :
/home/rusakov/wrkdir/dalton/Se/unc/tmp3:/usr/local/lib/dalton2013/dalton/basis

--- SEVERE ERROR, PROGRAM WILL BE ABORTED ---
Date and time (Linux) : Thu Nov 21 16:41:13 2013
Host name : NMR4

Reason: Non-existing basis set in HERBAS

>>>> Total CPU time used in DALTON: 29.57 seconds
>>>> Total wall time used in DALTON: 29.60 seconds


QTRACE dump of internal trace stack

========================
level module
========================
10 BASLIB
9 READ_MOL
8 READIN
7 HERMIT
6 GTNRGY
5 FNDGEO
4 RUNOPT
3 OPTMIN
2 DALTON
1 DALTON main
========================



The following directories have been searched :
/home/rusakov/wrkdir/dalton/Se/unc/tmp3:/usr/local/lib/dalton2013/dalton/basis

--- SEVERE ERROR, PROGRAM WILL BE ABORTED ---
Date and time (Linux) : Thu Nov 21 16:41:13 2013
Host name : NMR4

Reason: Non-existing basis set in HERBAS

________________________________________________________________________

It occurs only with ATOMBASIS option.

Re: Basis set label problem

Posted: 22 Nov 2013, 14:36
by bast
thank you for the bug report. i can confirm this.
we will look into the problem. a workaround in the meantime is to run without MPI -
in a sequential calculation this problem does not appear. something is broken
when transmitting basis set information to the worker processes.

Re: Basis set label problem

Posted: 30 Jan 2014, 13:02
by arnfinn
FYI, this has been fixed in Dalton2013.1:
http://daltonprogram.org/forum/viewtopic.php?f=17&t=100

Thanks for reporting it!