optomize error

ShaYANG

unread,

Jan 4, 2014, 3:03:29 AM1/4/14

to dirac-users

Dear all:
I want to calculate the structure of Tl(H2O)6 molecule with charge of plus three.I input two forms of input files,but both make mistakes.
Here I paste my 1and 2 input files and output files.
All the best!
Sha

1.inp

1.mol

1.out

2.inp

2.mol

2.out

Elham

unread,

Jan 4, 2014, 7:45:02 AM1/4/14

to dirac...@googlegroups.com

Hi ShaYANG
I'm not a Dirac Expert, but I'm a current user of the code and I guess that I know what is the problem.

1.you must be symmytry in mol.file because symmetry changed during in relaxation.

in 2.out put show that

Full group is: O(h)

Represented as: D2h

C 3 3 Z Y X A ---(D2h)

Cartesian coordinates xyz format (angstrom)

-------------------------------------------

19

Tl 0.0000000000 0.0000000000 0.0000000000

O 2.2729930000 0.0000000000 0.0000000000

O -2.2729930000 0.0000000000 0.0000000000

O 0.0000000000 0.0000000000 2.2729930000

O 0.0000000000 0.0000000000 -2.2729930000

O 0.0000000000 2.2729930000 0.0000000000

O 0.0000000000 -2.2729930000 0.0000000000

H 2.8480720000 0.0000000000 0.7694910000

H -2.8480720000 0.0000000000 0.7694910000

H 2.8480720000 0.0000000000 -0.7694910000

H -2.8480720000 0.0000000000 -0.7694910000

H 0.0000000000 0.7694910000 2.8480720000

H 0.0000000000 -0.7694910000 2.8480720000

H 0.0000000000 0.7694910000 -2.8480720000

H 0.0000000000 -0.7694910000 -2.8480720000

H 0.7694910000 2.8480720000 0.0000000000

H -0.7694910000 2.8480720000 0.0000000000

H 0.7694910000 -2.8480720000 0.0000000000

H -0.7694910000 -2.8480720000 0.0000000000

I same problem and dear Radovan this response for problem.

the problem is that the point group symmetry changes during the structure optimization.

in each structure optimization step DIRAC restarts from the converged DFCOEF

of the previous step. here it fails because the matrix dimensions are incompatible with

the DFCOEF of the previous step.

you can prevent this by explicitly forcing a specific point group symmetry:

http://www.diracprogram.org/doc/release-12/molecule_and_basis/molecule_using_mol.html#how-to-force-specific-symmetry

good luck!

radovan

I suggested search in mailing list and see same problem.

ShaYANG

unread,

Jan 5, 2014, 8:47:33 PM1/5/14

to dirac-users

Dear all:
I modified my input files as following:
(1)
.inp files:
**DIRAC
.TITLE
Tl(H2O)6 3+ ion
.WAVE FUNCTION
.OPTIMIZE
**INTEGRALS
*READIN
.UNCONTRACT
**HAMILTONIAN
.DOSSSS
**WAVE FUNCTION
.SCF
*END OF
.mol files:
DIRAC
Tl(H2O)6 3+ ion

C 3    3 3 Z Y X   A
      81.   1
Tl    0.000000 0.000000 0.000000
LARGE BASIS dyall.cv2z
      8.    6
O     0.000000 0.000000 2.272993
O     0.000000 2.272993 0.000000
O    -2.272993 0.000000 0.000000
O     0.000000 0.000000 -2.272993
O     0.000000 -2.272993 0.000000
O     2.272993 0.000000 0.000000
LARGE BASIS cc-pVDZ
      1.    12
H     0.000000 0.769491 2.848072
H    -0.769491 2.848072 0.000000
H    -2.848072 0.000000 0.769491
H     0.000000 -0.769491 -2.848072
H     0.769491 -2.848072 0.000000
H     2.848072 0.000000 -0.769491
H     0.000000 -0.769491 2.848072
H     0.769491 2.848072 0.000000
H    -2.848072 0.000000 -0.769491
H     0.000000 0.769491 -2.848072
H    -0.769491 -2.848072 0.000000
H     2.848072 0.000000 0.769491
LARGE BASIS cc-pVDZ
FINISH

(2)
.inp file:
**DIRAC
.TITLE
Tl(H2O)6 3+ ion
.WAVE FUNCTION
.OPTIMIZE
**INTEGRALS
*READIN
.UNCONTRACT
**HAMILTONIAN
.DOSSSS
**WAVE FUNCTION
.SCF
*SCF
.MAXITR
100
*END OF
.mol file:
DIRAC
Tl(H2O)6 3+ ion

C 3    3 3 Z Y X   A
      81.   1
Tl    0.000000 0.000000 0.000000
LARGE BASIS dyall.cv2z
      8.    3
O     0.000000 0.000000 2.272993
O     0.000000 2.272993 0.000000
O    -2.272993 0.000000 0.000000
LARGE BASIS cc-pVDZ
      1.    6
H     0.000000 0.769491 2.848072
H    -0.769491 2.848072 0.000000
H    -2.848072 0.000000 0.769491
H     0.000000 -0.769491 2.848072
H     0.769491 2.848072 0.000000
H    -2.848072 0.000000 -0.769491
LARGE BASIS cc-pVDZ
FINISH
These two input files both come out an error,
==== below this line is the stderr stream ====

Master node : --- SEVERE ERROR, PROGRAM WILL BE ABORTED ---
SYMGRP:IND.NE.MAXREP!
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
with errorcode -9999.

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.

How about this error and how to deal with it?
All the best !
            Sha

------------------ Original ------------------

From: "Elham";<e.chem...@gmail.com>;

Date: Sat, Jan 4, 2014 09:15 PM

To: "dirac-users"<dirac...@googlegroups.com>;

Subject: Re: optomize error

--
You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dirac-users...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Hans Jørgen Aagaard Jensen

unread,

Jan 6, 2014, 6:15:09 AM1/6/14

to dirac...@googlegroups.com

Dear Sha, please attach your output file.

___________________________________________

Hans Jørgen Aagaard Jensen

Department of Physics, Chemistry and Pharmacy
University of Southern Denmark
Odense, Denmark

ShaYANG

unread,

Jan 6, 2014, 9:47:41 AM1/6/14

to Hans Jørgen Aagaard Jensen

I attach an output file of the first input files.Another one has the same error!
Thank you!
Sha

------------------ Original ------------------

From: "Hans Jørgen Aagaard Jensen";<h...@sdu.dk>;

Date: Mon, Jan 6, 2014 07:15 PM

To: "dirac...@googlegroups.com"<dirac...@googlegroups.com>;

Subject: Re: optomize error

1_1.out

radovan bast

unread,

Jan 6, 2014, 10:09:42 AM1/6/14

to dirac...@googlegroups.com

dear Sha,

mind the fixed format.

you have set:

C___3__3_3__Z__Y__XA

but you probably wanted:

C_3____3_3_Z_Y_X___A

(the underscores are there so that the mail client does not move characters)

the other problem is that if you specify specific symmetry generators,

you should remove the symmetry-generated centers:

http://www.diracprogram.org/doc/release-13/molecule_and_basis/molecule_using_mol.html#how-to-force-specific-symmetry

i would also never include .DOSSSS right from the start. as soon as you have everything working

and you have a converged LVCORR structure you may want to switch on .DOSSSS.

good luck,

radovan

# PDC Center for High Performance Computing &

# Department of Theoretical Chemistry and Biology

# Royal Institute of Technology, Stockholm

# +46-8-790-6628

ShaYANG

unread,

Jan 6, 2014, 10:22:45 AM1/6/14

to dirac-users

Radovan:
Thank you! You mean the input files should be like this.

.inp files:
**DIRAC
.TITLE
Tl(H2O)6 3+ ion
.WAVE FUNCTION
.OPTIMIZE
**INTEGRALS
*READIN
.UNCONTRACT

**WAVE FUNCTION
.SCF
*END OF

.mol file:

C 3 3 3 Z Y XA

81. 1
Tl 0.000000 0.000000 0.000000
LARGE BASIS dyall.cv2z
8. 3
O 0.000000 0.000000 2.272993
O 0.000000 2.272993 0.000000
O -2.272993 0.000000 0.000000
LARGE BASIS cc-pVDZ
1. 6
H 0.000000 0.769491 2.848072
H -0.769491 2.848072 0.000000
H -2.848072 0.000000 0.769491
H 0.000000 -0.769491 2.848072
H 0.769491 2.848072 0.000000
H -2.848072 0.000000 -0.769491
LARGE BASIS cc-pVDZ
FINISH

Yes, I want to switch LVCORR structure on .DOSSSS. How should I do?
Sha!

------------------ Original ------------------

From: "radovan bast";<ba...@kth.se>;

Date: Mon, Jan 6, 2014 11:09 PM

ShaYANG

unread,

Jan 6, 2014, 10:56:29 AM1/6/14

to dirac-users

I modified the input file like you said, but there also a problem :Nuclei too close.

Why?
Thank you !
Sha

------------------ Original ------------------

From: "ShaYANG";<5502...@qq.com>;

Date: Mon, Jan 6, 2014 11:22 PM

1.inp

1.mol

1.out

radovan bast

unread,

Jan 6, 2014, 11:02:32 AM1/6/14

to dirac...@googlegroups.com

dear Sha,

please read this section i sent:

http://www.diracprogram.org/doc/release-13/molecule_and_basis/molecule_using_mol.html#how-to-force-specific-symmetry

the explanation for this error is there.

when you remove .DOSSSS, DIRAC defaults to LVCORR. please see:

http://www.diracprogram.org/doc/release-13/manual/hamiltonian.html?highlight=dossss#dossss

http://www.diracprogram.org/doc/release-13/manual/hamiltonian.html?highlight=lvcorr#lvcorr

best regards,

radovan

ShaYANG

unread,

Jan 6, 2014, 11:09:43 AM1/6/14

to dirac-users

Yes, I read the section of "How to force specific symmetry". I have deleted the coordinates of symmetry of three waters for the molecule of Tl(H2O)6 which has charge of 3 plus.But it also has an error of nuclei too close.

------------------ Original ------------------

From: "radovan bast";<ba...@kth.se>;

Date: Tue, Jan 7, 2014 00:02 AM

radovan bast

unread,

Jan 6, 2014, 11:12:27 AM1/6/14

to dirac...@googlegroups.com

this then means that there are still too many nuclei. you specify that you want 3 mirror planes

so you have to make sure that the 3 mirror planes generate your full molecule starting from the

unique centers.

Hans Jørgen Aagaard Jensen

unread,

Jan 6, 2014, 2:35:34 PM1/6/14

to <dirac-users@googlegroups.com>, dirac...@googlegroups.com

Of your six hydrogens in your last .mol file the last three are generated by symmetry operations on the first three, the last three should thus be removed from the .mol file (for example: the symmetry generated hydrogen from your first hydrogen after reflection in the xz-plane is on top of your fourth hydrogen, the symmetry generated hydrogen and the specified hydrogen no. 4 are therefore therefore "too close").

-- Hans Jørgen.

Sendt fra min iPad

Reply all

Reply to author

Forward