[Scipy-tickets] [SciPy] #1126: scipy.optimize.anneal does not respect lower/upper bounds

SciPy Trac scipy-tickets@scipy....
Thu Mar 4 09:28:19 CST 2010


#1126: scipy.optimize.anneal does not respect lower/upper bounds
--------------------------------+-------------------------------------------
 Reporter:  lboussouf           |       Owner:  somebody   
     Type:  defect              |      Status:  new        
 Priority:  normal              |   Milestone:  Unscheduled
Component:  scipy.optimize      |     Version:             
 Keywords:  bounds upper lower  |  
--------------------------------+-------------------------------------------

Comment(by lboussouf):

 Thank for the answer.

 But I don't agree about an expensive cost function outside the bounds. It
 is not a acceptable solution, and I hope nobody use that solution without
 knowing exactly what he does.

 About your conclusion, it is true : bounds are used only on the update.
 But it does not correctly implement the spirit of a bound constrained
 simulated annealing. In fact it is useless to constrain the update as it
 does not allow any control on the parameters.

 I propose to change this behaviour by a reprojection on the bounds. If
 scipy project tends to be used by mathematicians, it must propose correct
 algorithms. Let's start from that good code and correct it.

 But I am not familiar with the scipy project (maybe it is a good moment to
 start !). Can I modify this code to make it behave correctly and commit
 that modif to the repository or do I have to propose the modification in a
 forum and discuss it with other people (I am able to propose a
 modification of the code, but how to make it checked) ?

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/1126#comment:3>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list