[molpro-user] MS-CASPT2 geometry optimizations

[harding]

  I have been doing some geometry optimizations using the multi-state 
caspt2 method. It appears to me that unless I explicitly set 
variable=energy(1) it will not optimize on the ground electronic state 
but if I do set variable=energy(1) then it will not use analytic 
gradients. Subsequent frequency calculations do appear to be using 
analytic gradients and do appear to be using the ground electronic state 
without the need to set variable=energy(1) again. This happens in both 
the 2009 and 2010 versions.

  Am I doing something wrong? Any advice would be appreciated.

   Larry