[Scipy-tickets] [SciPy] #1644: scipy.optimize.fmin_bfgs not handling functions with boundaries that go infinite or very large correctly

SciPy Trac scipy-tickets@scipy....
Tue Apr 17 15:02:45 CDT 2012


#1644: scipy.optimize.fmin_bfgs not handling functions with boundaries that go
infinite  or very large correctly
----------------------------+-----------------------------------------------
 Reporter:  jsalvatier      |       Owner:  somebody   
     Type:  defect          |      Status:  new        
 Priority:  normal          |   Milestone:  Unscheduled
Component:  scipy.optimize  |     Version:  0.10.0     
 Keywords:                  |  
----------------------------+-----------------------------------------------
 Here's the toy problem I've set up:

 from scipy.optimize import fmin_bfgs, fmin_ncg
 from numpy import *
 import numpy as np

 def f(x ):
     if x < 0:
         return 1.79769313e+308
     else :
         return x + 1./x


 xs = fmin_bfgs(f, array( [10.]), retall = True)

 The solver returns [nan] as the solution.

 The problem is designed to be stiff: between 0 and 1, it slopes upward to
 infinity but between 1 and infinity, it slopes up at a slope of 1. Left of
 0 the function has a "nearly infinite" value. If bfgs encounters  a value
 that's larger than the current value, it should try a different step size,
 no?

 See this thread: http://mail.scipy.org/pipermail/scipy-
 user/2012-April/032088.html

 It seems that what was actually happening is that in line_search_wolfe2,
 calling scalar_search_wolfe2 calling _zoom calling _cubicmin was returning
 a NaN if you add a test at the end of _cubicmin testing for a NaN (and
 then returning None), it finds the right minimum. It looks like _quadmin
 probably would have the same problem.

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/1644>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list