[molpro-user] Too many operator domains during LCCSD calculation

[cong.wang]

Dear Everyone,

  Hello!

  I try to calculate C60 on DF-LCCSD/cc-pVTZ level, the end of the
output file said

*************************************************************************************
 Number of strong pairs:        1050
 Number of close pairs:          900     Rmin=      2.647
 Number of weak pairs:          2640     Rmin=      4.458
 Number of distant pairs:       2670     Rmin=      8.553

 Using   4 processors for pair domains. ntask=   7260  ngroup=   39
minbatch=    1  maxbatch=  256

 W-matrix paged, lenw=    100027980

 Average pair domain sizes:      100    (strong pairs: 123, close
pairs: 128, weak pairs: 104, distant pairs:  78)

 Smallest S-eigenvalue of domains:  0.1186E-05   Pair: 715  Symmetry: 1

 CPU-time for pair domains:            46.17 SEC

 Distance criteria for operator lists:     RMAXJ=      8.00  RMAXK=
  8.00  RMAXL=     15.00
 Distance criteria for operator domains:   RDOMJ=      0.00  RDOMK=      8.00
 Distance criterium for triples:           RMAXT=      3.00

 ERROR: TOO MANY OPERATOR DOMAINS (>****) PLEASE INCREASEMXDOM IN COMMON CLOCAL

 GA ERROR fehler on processor   0
********************************************************************

It is suggested increase mxdom in common clocal. What is "clocal"? I
cannot find this terminology in the manual...


Thank you very much
best regards
Cong Wang

Department of Chemistry
Laboratory for Instruction in Swedish
University of Helsinki
A.I. Virtanens plats 1
P.O. Box 55
FI-00014 University of Helsinki
FINLAND