[molpro-user] geometry optimization

[Borys Szefczyk]

Dear Molpro users,

I've been running optimization of a pretty big molecule, using
local CC2 method (although I don't think this is relevant here).
The optimization has apparently terminated succesfully, since
the output says:

 END OF GEOMETRY OPTIMIZATION.    TOTAL CPU:    940522.0 SEC

However, these are the final steps:

 ITER.   ENERGY(OLD)    ENERGY(NEW)      DE          GRADMAX     GRADNORM    GRADRMS     STEPMAX     STEPLEN     STEPRMS
(...)
  11  -871.10443499  -871.10443839    -0.00000339  0.00007316  0.00020166  0.00000129  0.03971752  0.04648610  0.00632596
  12  -871.10443839  -871.10444174    -0.00000335  0.00006415  0.00019509  0.00000125  0.03929715  0.04593770  0.00625133
  13  -871.10444174  -871.10444497    -0.00000323  0.00007093  0.00019072  0.00000122  0.03892657  0.04551436  0.00619372

The standard criteria for convergence are GRADMAX < 3e-4 & DE < 1e-6 (not satisfied)
or
GRADMAX < 3e-4 & STEPMAX < 3e-4  - not satisfied, unless I am mistaken...
The manual says "the maximum component of the step to be less than 3 * 10^-4",
this is STEPMAX the manual is talking about, isn't it?

So here is my dumb question: did my optimization converge or not?

Regards,
Borys Szefczyk

-- 
  Molecular Modelling & Quantum Chemistry Group,
  Institute of Physical & Theoretical Chemistry,
  Wroclaw University of Technology
  http://ichfit.ch.pwr.wroc.pl/people/szefczyk