I agree the input of $transt is a bit ugly.
We'll see what can be done to simplify it.
As to CI on K-S DFT reference, it is possible. You need to add
$contrl wide=.t. $end
to your dft energy run, then pick up the complete $vec group
from the punch file, add it to the input file, and in addition add:
$contrl scftyp=none dfttyp=none cityp=guga (or aldet) $end $guess asis=.t. guess=moread norb= put here the total number of MOs
and then run the second job.
Hope this helps.
On Tue Nov 3 '15 8:41pm, PeterP wrote
>thank you for the help and the examples. I managed to set up my calculations correclty and get all the requested informations. Indeed one can't get these data from an prior RUNTYPE=ENERGY job but one need's to use RUNTYPE=TRANSITN. The input-file for my Nitrobenze example than needs to look like this:
> $CONTRL SCFTYP=NONE RUNTYP=TRANSITN EXETYP=RUN MAXIT=30 MULT=1 $END
> $CONTRL CITYP=GUGA $END
> $SYSTEM MWORDS=20 $END
> $BASIS GBASIS=N31 NGAUSS=6 NDFUNC=1 NPFUNC=1 $END
> $GUESS GUESS=MOREAD NORB=160 $END
> $GUGDIA NSTATE=5 $END
> $CIDRT1 GROUP=C2V ISTSYM=0 NFZC=9 NDOC=23 NVAL=128 IEXCIT=2 $END
> $TRANST NUMVEC=1 NUMCI=1 NOCC=160 NFZC=160 IROOTS(1)=3 NSTATE(1)=5 $END
> $SCF DIRSCF=.TRUE. DIIS=.T. $END
>So all necessary commands are in $CIRDRT - which would be the same for runtype energy - and $TRANST. Here one needs to set NUMVEC=NUMCI=1 and NOCC=NFZC=(NFZC+NDOC+NVAL). IROOT and NSTATE are arrays. Actually I don't understand why NOCC must equal NFZC in $TRANST ... but it works.
>But still: any ideas on the KS-DFT reference ?