Molpro: xyz input and atom reordering

[Alex Brown]

Dear All,

  When running Molpro using Cartesian coordinates, I have noticed that
Molpro often reorders the atoms (see simple examples below). Since for my
applications (or at least the scripts I use to search output), I would
like my input ordering to conform to the Molpro atom ordering, can anyone
tell me how Molpro performs the reordering? That way I can make sure that
my input atom ordering will conform EXACTLY to my output ordering. Note
that there does not seem to be an option to stop this from happening.
Also, I realise that I could modify my script to double-check (and
correct) for re-ordering but I'm curious as to how (and why) Molpro
reorders the atoms, and if input = output, it will be simpler.

Example One Input:
geomtyp=xyz
 geometry={
 3
! number of atoms
 This is an example of geometry input for water with an XYZ file
 O ,0.0000000000,0.0000000000,-0.1302052882
 H ,1.4891244004,0.0000000000, 1.0332262019
 H,-1.4891244004,0.0000000000, 1.0332262019
 }
 hf

Ordering of Atoms by Molpro (taken from output file and input ordering =
Molpro ordering):
 ATOMIC COORDINATES

 NR  ATOM    CHARGE       X              Y              Z

   1  O       8.00    0.000000000    0.000000000   -0.246052317
   2  H       1.00    2.814037080    0.000000000    1.952514406
   3  H       1.00   -2.814037080    0.000000000    1.952514406


Example Two input (switch O and H1 order):
 geomtyp=xyz
 geometry={
 3
! number of atoms
 This is an example of geometry input for water with an XYZ file
 H ,1.4891244004,0.0000000000, 1.0332262019
 O ,0.0000000000,0.0000000000,-0.1302052882
 H,-1.4891244004,0.0000000000, 1.0332262019
 }
 hf

Example Two Atom ordering in Molpro:
 ATOMIC COORDINATES

 NR  ATOM    CHARGE       X              Y              Z

   1  H       1.00    2.814037080    0.000000000    1.952514406
   2  H       1.00   -2.814037080    0.000000000    1.952514406
   3  O       8.00    0.000000000    0.000000000   -0.246052317

So, it does not conform exactly to the input (H2 and O have been
switched). Any ideas?

Thanks,
Al

Dr. A. Brown
Post-doctoral Research Associate
Department of Chemistry
Atwood Hall
1515 Pierce Drive
Emory University
Atlanta, GA
30322
http://userwww.service.emory.edu/~abrown8/