Error Exit - linearly dependent basis/illegal negative file offset

[Dave Moore]

Hi,

	I have had many similar messages in the past .. and, with one exception,
they were always due to an improperly constructed basis set .. especially
when dummy atoms were used.  Be very careful about your input symmetry when
doing geometry optimizations ... if an optimization move breaks symmetry,
then you will get an error, which is sometimes (but not always) the "BASIS
LINEARLY DEPENDENT OR WRONG S" error.

	The other time I have gotten the error is for single point calculations on
systems with large distances between atoms (6.0 Angstroms are more -- for
example in a PES scan) .. and in this case I think the error is due to a bug
in the symmetry determination routine of the program .. anyway, I was always
able to work around it by rewriting my Z-matrix .. although it was not
always obvious how to do it, so sometimes it took quite a long time to get
one that would work.

Hope this helps,

Dave Moore


-----Original Message-----
From: owner-molpro-user at molpro.bham.ac.uk
[mailto:owner-molpro-user at molpro.bham.ac.uk]On Behalf Of Aiko Huckauf
Sent: Thursday, August 22, 2002 11:01 PM
To: molpro-user at molpro.net
Subject: Error Exit - linearly dependent basis/illegal negative file
offset


Dear Sir or Madam,

in my mp2 and ccsd(t) calculations Molpro sometimes stopped with
error messages that do not really help me to understand why.

I have performed a test run using the same input data, varying only
method and/or basis set, with the following results:

--

MP2 cc-pVDZ       works fine
MP2 cc-pVTZ       works fine
MP2 cc-pVQZ       f_seek: illegal negative file offset on file
                  T1100017125.TMP, unit 11, offset=-2147479552

MP2 aug-cc-pVTZ   ERROR: BASIS LINEARLY DEPENDENT OR WRONG S
MP2 aug-cc-pVQZ   ERROR: BASIS LINEARLY DEPENDENT OR WRONG S

CCSD(T) cc-pVDZ       works fine
CCSD(T) cc-pVQZ       f_seek: illegal negative file offset on file
                      T1100017166.TMP, unit 11, offset=-2147479552

CCSD(T) aug-cc-pVTZ   ERROR: BASIS LINEARLY DEPENDENT OR WRONG S
CCSD(T) aug-cc-pVQZ   ERROR: BASIS LINEARLY DEPENDENT OR WRONG S

--

I am using Molpro 2000.1 on a RedHat Linux 7.0 system (kernel 2.4.18,
AMD Athlon 1333 MHz, 1.5 GB RAM).

I would appreciate very much if somebody could give me a hint at the
causes of the above-mentioned errors (and help me to avoid them).

Thank you very much in advance!

Aiko Huckauf

--
Dr. Aiko Huckauf                       Department of Chemistry
Fon: 780-492-4635                      University of Alberta
Fax: 780-492-8231                      Edmonton AB  T6G 2G2
E-Mail: Aiko.Huckauf at ualberta.ca       Canada