[SciPy-User] using fmin_tnc

[Jason Rennie]

On Wed, May 11, 2011 at 12:24 PM, Joon Ro <joonpyro at gmail.com> wrote:

> So the documentation does say x0 is a list of floats. I haven't tried it
> but it should work with a one dimensional vector as well.  I'm not sure if
> the routine passes x as a list or not to the objective function though.
>

I have tried it, and, yes, it accepts a numpy array.  As Joon noted, you
need to flatten your parameters into an array.  Consider creating functions
for converting between the two formats (two matrices, single vector).  I
typically use a class instance to keep track of the specifications for the
conversion, such as the sizes of the two matrices.  l_bfgs_b has a similar
interface and has very smart convergence criterion AFAICT, so if you have
trouble with tnc, try l_bfgs_b (even if your problem is unconstrained).

Cheers,

Jason

-- 
Jason Rennie
Research Scientist
Google/ITA Software
617-714-2645
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20110511/71fc50c4/attachment.html>