[SciPy-user] odeint

LOPEZ GARCIA DE LOMANA, ADRIAN alopez at imim.es
Wed Oct 5 07:16:17 CDT 2005


I really agree with you, Robert. The result of the simulation is correct, but the amount of the error is so big that I guess I'm loosing performance on the speed of the program. Do you have any suggestion about how to silence these errors?

Thanks, 

Adrián.

-----Original Message-----
From: scipy-user-bounces at scipy.net on behalf of Robert Kern
Sent: Wed 05/10/2005 12:33
To: SciPy Users List
Subject: Re: [SciPy-user] odeint
 
LOPEZ GARCIA DE LOMANA, ADRIAN wrote:
> Thanks for your answer, I'll check if it works in my machine. I guess your're right, the errors appear when the values of the simulation are close to zero.
> 
> The problem is that I'm not just interested on solving this specific example but it is part of a bigger program where I search for the parameters, run the simulations, and evaluate a score given some data. I'll check how if I can integrate your suggestion into the program.
> 
> But, don't you think we are talking about a bug? Overall, if we want to compare scipy with Octave ... I mean, I definetly prefer SciPy, it's much faster, but I also want it to be reliable, and I don't want to be caring if the simulation is "too" close to zero.

What exactly is the bug? odeint can't know that some of your parameters
aren't supposed to be negative and thus automatically avoid them. When
you ran your simulation, you got some output warning you that the
function was raising exceptions, but the simulation completed did it
not? odeint correctly determined that its guesses overshot the boundary
on the domain of valid parameters, backed up and continued integrating.
It might be worth adding an argument to the function to silence the
output of such error messages.

-- 
Robert Kern
rkern at ucsd.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter

_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.net
http://www.scipy.net/mailman/listinfo/scipy-user



More information about the SciPy-user mailing list