[SciPy-User] Setting an absolute tolerance to fmin_l_bfgs_b

[Jose Gomez-Dans]

Hi,
I'm minimising a function with fmin_l_bfgs_b. The function has many local
minima, so I need to have a low high tolerance (factr parameter) so as not
to get trapped in one. Also, the area around the global minimum is usually
quite flat, and I have an expectation of the value of the function at the
minimum (although not of where the minimum is!). When the algorithm reaches
the minimum, it goes around for a long time, optimising the function further
and further. This is overkill, as I have ways to calculate the uncertainty
in the parameters post-optimisation and I would rather the optimisation
stopped once the function is under a given threshold. In the fortran
version, I just have an if statement and bail out of it, but I was wondering
whether the scipy version has something similar (the docs imply that you can
only tweak m, pgtol and factr). Any ideas?


Thanks!
Jose
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20110623/dd0047af/attachment.html>