[Jmol-developers] CDK RFC #23: Storage of geometry optimization results

[E.L. W. <eg...@sc...>]

RFC #23
 
Name:        Storage of geometry optimization results
Date:          2003-09-04

Proposal

Geometry optimization results should be stored as a ChemFile with three 
ChemSequences; the first with only one ChemModel being the structure given to 
the optimization program, the second containing the complete optimization 
run, and the last ChemSequence again with only one ChemModel being the
optimized structure.

Reason

In the last few weeks I've ported two Gaussian readers from Jmol to CDK (98 
and 03), which are both used by Jmol-HEAD... All gaussian 'frames' are now 
stored in one ChemSequence, reflecting the behavior of the Jmol versions of 
these readers... However, the .log file contains not just the geometry 
optimization run, but also the starting material... This would suggest the 
put the later in a separate ChemSequence... And what about the optimized 
structure which should be used to put in a XYZ file when it is converted into 
that format... Thus, we might need a separate ChemSequence with a duplicate 
of the optimzed geometry too, hence the proposal. The problem is applicable
to other geometry optimization programs too, like ABINIT.

Egon