PC GAMESS/Firefly-related discussion club

Learn how to ask questions correctly

Alex Granovsky
gran@classic.chem.msu.su

this problem is caused by integer overflow.
Namely, the number of CSF generated by DRT code
is huge and causes variable used to count CSFs
to become negative. Later, this (wrong) value
is used to calculate addresses of working arrays
and (because the counter is weird) these arrays
fall out of the region of allocated memory, causing
Firefly to trap. We'll add check for overflow
condition to this code to fix this bug.
However, there should not be any problems with GUGA
code in the absence of overflow.  

Anyway, as the number of CSFs is huge it wouldn't be
possible to run this job. However, you can use smaller
active space.

Note, your job is not of FOCI type, as you did not provide
NEXT in the $cidrt group!

Note also that for true FOCI/SOCI the CASTRF=.TRUE. option
would require lots of memory and is not always the fastest one!

There is another option you can try, which is specific to Firefly.
Namely, you can use the following sample $CIDRT group

 $CIDRT GROUP=C1 FORS=.FALSE. FOCI=.TRUE. SOCI=.FALSE. NFZC=7 NDOC=18 NVAL=6 next=-1 iexcit=3 $END

For example, the $CIDRT above generates 10234375 CSFs, which is
a tractable number. Although this CI space is not the full FOCI
space, it is a rather good approximation to full FOCI.

Finally, you MUST use any existing abelian symmetry to further
reduce the number of CSFs!

Regards,
Alex Granovsky






On Thu Mar 19 '09 1:26pm, Veinardi Suendo wrote
-----------------------------------------------
>Dear Prof. Alex and Colleagues,

>I got an error message during GUGA CI calculation:

>  THE WAVEFUNCTION CONTAINS-889688496The instruction at 0x00547b1e referenced memory at 0x30e82afc.
>The memory could not be written.

>Is it a bug or maybe I made mistake in the input commands? My input commands are:

>!   File created by MacMolPlt 7.2
> $CONTRL SCFTYP=RHF RUNTYP=ENERGY CITYP=GUGA INTTYP=HONDO
>  EXETYP=CHECK MAXIT=50 MULT=1 ICUT=11 ITOL=30
>  FSTINT=.TRUE. GENCON=.TRUE. $END
> $SYSTEM TIMLIM=525600 MEMORY=4000000 KDIAG=0 NOJAC=100
>  FASTF=.TRUE. $END
> $P2P p2p=.t. dlb=.t. xdlb=.t. mxbuf=2048 $END
> $SMP Call64=.t. httpar=.t. httalt=.t. $END
> $BASIS GBASIS=N31 NGAUSS=6 NDFUNC=1 DIFFSP=.TRUE. DIFFS=.TRUE. $END
> $GUESS GUESS=MOREAD NORB=25 KDIAG=-2 PRTMO=.TRUE. $END
>$SCF DIRSCF=.TRUE. FDIFF=.FALSE. NCONV=6
>  EXTRAP=.TRUE. DAMP=.TRUE. SHIFT=.TRUE. RSTRCT=.TRUE.
>  DIIS=.TRUE. SOSCF=.FALSE. $END
> $CIDRT GROUP=C1 FORS=.FALSE. FOCI=.TRUE. SOCI=.FALSE. NFZC=7
>  NDOC=18 NVAL=6 $END
> $TRANS MPTRAN=2 DIRTRF=.TRUE. AOINTS=DIST ALTPAR=.TRUE. MODE=112 $END
> $CIINP CASTRF=.TRUE. $END
> $GUGEM PACK2=.TRUE. $END
> $DATA
>Phenol 6-31++G(d) GUGA
>C1
>C     6.0    -5.15420     0.69284     0.00000
>C     6.0    -5.14472    -0.69619     0.00000
>C     6.0    -3.94883    -1.39409     0.00000
>...

>I do hope it is not a bug, because we are running some other calculations right now. Please give us some advices, if we made mistakes in our input file. Thank you very much in advance.

>Best regards,

>Veinardi

[ This message was edited on Fri Mar 20 '09 at 10:54pm by the author ]