[molpro-user] Data from different molecule after restart

[Sabrina Kemmerer]

Hello everyone,
I am trying to do a transition state optimization. Everything went fine 
so far until my calculation reached the time limit of the computing 
cluster. I tried to restart the calculation (I just started again with 
the same input.), but then got this error:
?ERROR IN GET_GUESS: RECORD    2100.1 CONTAINS DATA FOR DIFFERENT 
MOLECULE:   5   5
I am sure I did not touch any of the files 1,2 or 3. Has anyone an idea, 
why this error occurs?

This is the calculation I am trying to do:

***
MEMORY,1000,M
FILE,1,cclbr-fhpts.int
FILE,2,cclbr-fhpts.wfu
FILE,3,cclbr-fhpts.aux

ANGSTROM
GEOMETRY={
C
H,1,1.248
F,1,1.753,2,46.1
Br,1,1.753,3,110,2,180,0
Cl,1,1.626,4,128.7,2,170,0
}
PUT,MOLDEN,cclbr-fhptsstart.molden

SET,CHARGE=1

BASIS={
ECP,Br,ECP10MDF
default,cc-pVTZ-F12
INCLUDE,cc-pVTZ-PP-F12

set,df,CONTEXT=MP2FIT
DEFAULT=aug-cc-pwCVTZ
INCLUDE,cc-pVTZ-PP-F12_MP2FIT

set,jk,CONTEXT=JKFIT
DEFAULT=cc-pVTZ
Br=def2-QZVPP

set,ri,CONTEXT=OPTRI
DEFAULT=cc-pVTZ-F12
INCLUDE,cc-pVTZ-PP-F12_OPTRI
}

RHF
{UCCSD(T)-F12b,DF_BASIS=df,DF_BASIS_EXCH=jk,RI_BASIS=ri,GEM_BETA=[1.4];MAXIT,100}
{OPTG,ROOT=2,METHOD=QSD}
PUT,MOLDEN,cclbr-fhpts.molden
{FREQUENCIES;PRINT,HESSIAN}
---