Conjugate gradients minimizer

[Andrew Walkingshaw]

Hi. Ages since I last posted to c.l.p, and this is a long shot, but here
goes.

Does anyone have a conjugate-gradients minimization routine (failing
that, even a line-search-using-gradients algorithm would be extremely
useful):

a) in "pure" python (NumPy is *just about* OK for my purposes; C
extensions aren't);

b) under the Python/2-clause BSD/X11/public-domain/similar license.

(this is as part of work I'm doing into a problem in materials physics,
and I intend to release the code I'm working on under a Free Software
license at some point[1] - but depending on your perception, it could be
construed to be linking to non-free software, and I have neither the time
or inclination to get into a license war right now. Commercial
exploitation, however, is right out.)

... which I could either adapt or even just *read*? I've looked at
Numerical Methods in C, and though it makes the algorithms pretty clear,
it's so unPythonic it sets my teeth on edge. 

Many thanks,
Andrew

(The subject email address *is* real, but is read infrequently. Trim
the obvious extension for quicker replies...)

[1] http://www.lexical.org.uk/created/anarcryst-latest.tar.gz is some
of my other code: it won't be interesting unless you really need
autogenerated disordered crystalline structures, but it *is* GPL. :)

-- 
Andrew Walkingshaw | andrew-usenet at lexical.org.uk