[molpro-user] integral thresholds

[Ben Mintz]

I am trying to change the threshold for the neglect of two-electron 
integrals.  The manual says that this can be done with the following 
"gthresh,twoint=1.d-n".  When I do this I get the following in my 
output.  It says that the two-electron integral threshold is changed to 
1.d-6, but when the integrals are computed the output says, "Contracted 
2-electron integrals neglected if value below      1.0D-11".  Is gthresh 
the correct way to change the threshold for the two-electron integrals, 
or is there another way?  If I am doing this correctly, why does the 
calculation say that the threshold is changed in one part of the output 
but using a different threshold in another part?

 THRESHOLDS:

 ZERO    =  1.00D-12  ONEINT  =  1.00D-12  TWOINT  =  1.00D-06  PREFAC  
=  1.00D-14  LOCALI  =  1.00D-09  EORDER  =  1.00D-04
 ENERGY  =  1.00D-06  GRADIENT=  1.00D-02  STEP    =  1.00D-03  ORBITAL 
=  1.00D-05  CIVEC   =  1.00D-05  COEFF   =  1.00D-04
 PRINTCI =  5.00D-02  PUNCHCI =  9.90D+01  OPTGRAD =  3.00D-04  
OPTENERG=  1.00D-06  OPTSTEP =  3.00D-04  THRGRAD =  1.00D-09
 COMPRESS=  1.00D-11  VARMIN  =  1.00D-07


 Variable memory set to   30000000 words,  buffer space   230000 words

 **********************************************************************************************************************************

**********************************************************************************************************************************
 Contracted 2-electron integrals neglected if value below      1.0D-11
 AO integral compression algorithm  1   Integral accuracy      1.0D-11

     14.156 MB (compressed) written to integral file ( 44.5%)

     Node minimum: 14.156 MB, node maximum: 14.156 MB


 NUMBER OF SORTED TWO-ELECTRON INTEGRALS:    3751849.     BUFFER 
LENGTH:  32768
 NUMBER OF SEGMENTS:   1  SEGMENT LENGTH:    3751849      RECORD LENGTH: 
524288