A test you could run is trying the set of thetas = [170.00, 170.01, 170.02, 170.03]. In some cases, if your convergence criteria are not
tight enough, you may see your energies drop just by
performing more calculations. I'm not sure, but I believe the
initial guess used in performing the second (or higher) iteration of your do loop is not the same initial guess in the first iteration of your do loop.
A simple test to see if this is the problem you are encountering is to run each calculation in a different file instead of using a do loop. If you get different answers, then I think increased convergence criteria is necessary.
Cheers, Tim Giese
University of Minnesota [EMAIL PROTECTED]
