[CP2K:1685] Re: PM6 - MTD - Matrix diagonalization failed

Laino Teodoro teodor... at gmail.com
Wed Dec 24 16:04:22 UTC 2008


Hi Francesco,

that was a perfect job. it took only 5 minutes to find out that there  
were no particular errors in your setup.
What you observe is something numerical connected to the  
diagonalization..

I wanted to be sure as well that your geometry had no particular  
problems.

So the only solution to override your problem is switching to &OT for  
the SCF:

     &SCF
       SCF_GUESS MOPAC
       EPS_SCF 1.0E-06
       &OT
         PRECONDITIONER FULL_SINGLE_INVERSE
         MINIMIZER DIIS
         N_DIIS 7
       &END
       &PRINT
         &RESTART OFF
         &END RESTART
       &END PRINT
     &END SCF

The other suggestion I have is to use analytical_gradients. The run  
will be much faster.
Hope this helps!
Have nice holidays!
Teo

On 24 Dec 2008, at 16:34, Francesco Ragone wrote:

> Hi Teo, Hi to everybody
>
> above all I wish to the cp2k google group good festivities ad a  
> GREAT 2009.
>
>
>
> I think I've done all what you asked for.
> I've redone the jobs three times to be more sure  I'm not doing  
> errors.
>
> In this case I've used a cp2k.sopt compiled on December 18th
> on a intel xeon quad core
> fortran compiler = ifort 10.0.0.25
> (but I get the problem also with a parallel version compiled on  
> July 20th 2008).
>
>
> I've attached a tar.gz archive in this message.
> In the tar you will find every file you need (input. restart and  
> the little output of a restarted job that shows the error I  
> obtained). The job runs just for few seconds and stops everytime at  
> the step 83819.
>
> Here the arch file I've used to compile the cp2k.sopt.
>
> ###############
> #  arch file sopt
> ###############
> INTEL_INC= /home/programs/intel/64/fce/10.0.025/include/
> CC       = cc
> CPP      =
> FC       = ifort
> LD       = ifort
> AR       = ar -r
> DFLAGS   = -D__INTEL -D__FFTSG -D__FFTW3
> CPPFLAGS = -C -traditional $(DFLAGS) -I$(INTEL_INC) -I$(FFTW3_INC)
> FCFLAGS  = $(DFLAGS) -I$(INTEL_INC) -I$(FFTW3_INC) -O2 -xW -heap- 
> arrays 64 -fpp -free
> LDFLAGS  = $(FCFLAGS)
>
> LIBS     = -L$(MKL_LIB) $(MKL_LIB)/libmkl_lapack.a   $(MKL_LIB)/ 
> libmkl_em64t.a $(MKL_LIB)/libguide.a -lpthread $(FFTW3_LIB)/libfftw3.a
>
> OBJECTS_ARCHITECTURE = machine_intel.o
>
> ####################################
>
> Thanks for all,
> Francesco Ragone
>
>
>
> 2008/12/19 Teodoro Laino <teodor... at gmail.com>
>
> Is there any reason why you didn't put somewhere your restart file?
> Let's iterate once more.. now with detailed instructions..
> 0) Take your input and your restart.. First (step 1-4) check that the
> file you send can EXACTLY reproduce the same error message you  
> claim to
> see.
> 1) Create a new directory..
> 2) put there the input and the restart file and all other files needed
> to run CP2K
> 3) run CP2K in the newly created directory (modify the input in  
> order to
> have the path pointing to this local directory!)
> 4) Is the error reproducible ?
>     YES) Great! clean the directory from possible files created by  
> your
> very last run.
>               Archive the whole directory and GZIP it.
>     NO) That's bad.. something is wrong.. go back to 0 checking that
> you are really doing something meaningful.
>
> 5) Very good! You reached already point Nr.5! What should you do now
> with the .tar.gz ?
>     Two options:
>      -) Attach it to your message (ok.. if you use the browser to read
> google groups you can't attach file.. if you use a standard mail  
> program
> like pine or gmail or thunderbird or OS/X Mail or Outlook you CAN!)
>      -) Load the gzipped archive in the File section of the cp2k  
> google
> group.
> 6) Send an acknowledgment message saying where is your archived  
> directory.
>
> Teo
>
> Francesco wrote:
> > I'm sorry, I wrongly though sending all the input could be more
> > "wasting time"...
> >
> > Here the input
> > &EXT_RESTART
> >    RESTART_FILE_NAME cp2k-1.restart
> > &END EXT_RESTART
> > &MOTION
> >   &MD
> >     COMVEL_TOL 1E-08
> >     ANGVEL_ZERO T
> >     ENSEMBLE NVT
> >     STEPS 100000
> >     TIMESTEP 0.5
> >     TEMPERATURE 300.0
> >     &THERMOSTAT
> >        TYPE NOSE
> >        REGION GLOBAL
> >        &NOSE
> >          LENGTH 4
> >          TIMECON 10
> >        &END NOSE
> >     &END THERMOSTAT
> >   &END MD
> >   &FREE_ENERGY
> >     METHOD METADYN
> >     &METADYN
> >       DO_HILLS T
> >       LAGRANGE T
> >       WW 0.002
> >       TEMPERATURE 300.0
> >       NT_HILLS 150
> >       &METAVAR
> >         MASS 10.0
> >         WIDTH 0.15
> >         LAMBDA 0.3
> >         COLVAR 1
> >       &END METAVAR
> >       &METAVAR
> >         MASS 10.0
> >         WIDTH 0.15
> >         LAMBDA 0.3
> >         COLVAR 2
> >       &END METAVAR
> >       &PRINT
> >         &COLVAR
> >           FILENAME =colvar
> >         &END COLVAR
> >       &END PRINT
> >     &END METADYN
> >   &END FREE_ENERGY
> >   &PRINT
> >     &TRAJECTORY
> >       &EACH
> >         MD 20
> >       &END EACH
> >     &END TRAJECTORY
> >     &VELOCITIES
> >       &EACH
> >         MD 20
> >       &END EACH
> >     &END VELOCITIES
> >     &RESTART
> >       &EACH
> >         MD 200
> >       &END EACH
> >     &END RESTART
> >   &END PRINT
> >   &GEO_OPT
> >   &END GEO_OPT
> > &END MOTION
> > &FORCE_EVAL
> >   METHOD QS
> >   &DFT
> >      &POISSON
> >        POISSON_SOLVER WAVELET
> >        PERIODIC NONE
> >      &END POISSON
> >     &QS
> >       METHOD PM6
> >       &SE
> >         ANALYTICAL_GRADIENTS F
> >         STO_NG 6
> >       &END SE
> >     &END QS
> >     &SCF
> >       SCF_GUESS MOPAC
> >       EPS_SCF 1.0E-06
> >       &PRINT
> >         &RESTART OFF
> >         &END RESTART
> >       &END PRINT
> >     &END SCF
> >   &END DFT
> >   &SUBSYS
> >     &TOPOLOGY
> >        CENTER_COORDINATES T
> >     &END TOPOLOGY
> >     &CELL
> >       ABC 26.0 26.0 26.0
> >       PERIODIC NONE
> >     &END CELL
> >     &COORD
> >     &END COORD
> >     &VELOCITY
> >     &END VELOCITY
> >     &COLVAR
> >       &DISTANCE
> >          ATOMS 27 26
> >       &END DISTANCE
> >     &END COLVAR
> >     &COLVAR
> >       &DISTANCE
> >          ATOMS 27 28
> >       &END DISTANCE
> >     &END COLVAR
> >   &END SUBSYS
> > &END FORCE_EVAL
> > &GLOBAL
> >   PROJECT_NAME cp2k
> >   RUN_TYPE MD
> >   PRINT_LEVEL LOW
> > &END GLOBAL
> >
> >
> > The starting geometry is below, but clearly this input takes the
> > geometry into the restart
> >
> >   C     -3.904899    -0.845867    -0.830939
> >   C     -3.115406    -0.890307     0.340835
> >   C     -3.599451    -0.396334     1.574198
> >   C     -4.887541     0.183271     1.591096
> >   C     -5.680774     0.296956     0.432897
> >   C     -5.169421    -0.219199    -0.765951
> >   N     -1.846646    -1.569609     0.286082
> >   C     -0.611450    -1.029407     0.125897
> >   N      0.335642    -1.985359     0.351002
> >   C     -0.276835    -3.263182     0.758183
> >   C     -1.773350    -2.980331     0.698344
> >   Ru    -0.339300     0.882278    -0.430166
> >   Cl    -0.282438    -0.254470    -2.628559
> >   C      1.776854    -1.939823     0.269209
> >   C      2.412861    -2.373622    -0.928990
> >   C      3.820023    -2.491550    -0.925595
> >   C      4.589296    -2.256854     0.224169
> >   C      3.917252    -1.869031     1.397277
> >   C      2.521229    -1.720698     1.456678
> >   C      1.629186    -2.691156    -2.186777
> >   C      6.096313    -2.431016     0.215235
> >   C      1.859430    -1.404193     2.775865
> >   C     -2.777112    -0.420383     2.831776
> >   C     -7.031509     0.964005     0.485639
> >   C     -3.454780    -1.508923    -2.106033
> >   C      1.120594     2.015943    -1.309491
> >   C     -0.116796     2.948894    -1.432620
> >   C     -1.489295     2.182081    -1.404068
> >   C      2.290728     2.468518    -0.512745
> >   C      3.449983     1.654335    -0.519204
> >   C      4.609490     2.043789     0.167814
> >   C      4.641980     3.255113     0.872610
> >   C      3.509486     4.084132     0.872232
> >   C      2.348172     3.700326     0.179906
> >   Cl    -0.517640     1.922871     1.679461
> >   H      1.413717     1.582114    -2.293686
> >   H      3.433486     0.701297    -1.068882
> >   H      5.508413     1.395315     0.136859
> >   H      5.543403     3.551922     1.418539
> >   H      3.529550     5.046502     1.410689
> >   H      1.489871     4.379851     0.167420
> >   H     -2.307121    -3.614375    -0.029110
> >   H      0.046893    -4.066343     0.075271
> >   H     -2.273125     2.696615    -0.823622
> >   H     -1.821294     1.827685    -2.396489
> >   H     -0.080799     3.467930    -2.414564
> >   H     -0.102597     3.703587    -0.626201
> >   H      4.325133    -2.804227    -1.857099
> >   H      4.497666    -1.714829     2.329436
> >   H     -5.272016     0.571138     2.545068
> >   H     -5.771895    -0.146649    -1.689732
> >   H     -2.308490    -1.408118     3.020504
> >   H     -1.953357     0.319217     2.766190
> >   H     -3.405589    -0.167401     3.713717
> >   H     -7.639595     0.722379    -0.408070
> >   H     -7.597934     0.653831     1.384041
> >   H     -6.930373     2.075036     0.533473
> >   H     -4.020510    -1.114678    -2.969242
> >   H     -2.373731    -1.364006    -2.302804
> >   H     -3.653448    -2.595613    -2.061599
> >   H      1.252096    -2.268281     3.137155
> >   H      2.615839    -1.182339     3.556384
> >   H      1.188145    -0.532863     2.710485
> >   H      0.997535    -1.831859    -2.492754
> >   H      2.305975    -2.949553    -3.015588
> >   H      0.940929    -3.553412    -2.045529
> >   H      6.622210    -1.481323     0.447628
> >   H      6.421685    -3.163141     0.980358
> >   H      6.463369    -2.784327    -0.765285
> >   H     -2.274158    -3.108035     1.673925
> >   H      0.072158    -3.534256     1.770801
> > >
> >
>
>
>
>
>
> >
> <TRY.tar.gz>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20081224/1a66e6f2/attachment.htm>


More information about the CP2K-user mailing list