[SciPy-User] optimize.fmin_cg and optimize.fmin_bgfs optimize to 0

[Pauli Virtanen]

Tue, 07 Jun 2011 15:23:46 -0400, Patrick Holvey wrote:
[clip]
> However, when
> I use the gradient in the optimization, all of the atom positions shoot
> right to the origin (so they're all at 0,0,0) after just 2 function
> calls and 1 gradient call, which seems very odd to me.  So I tried
> fmin_bgfs with the gradient and the same thing happened.  Does anyone
> have any experience with analytic gradients where this has happened to
> them?  I'm confused as to whether the problem is in my gradient
> implementation or in how I'm passing the gradient or what.

Your Box.Forces(self, xyz) method modifies the input `xyz` argument.
This you should not do --- the optimizer expects that you do not alter
the current position this way.

Try replacing

	vectorfield=xyz

with

	vectorfield = numpy.zeros_like(xyz)

or put

	xyz = xyz.copy()

in the beginning of the routine.