lsode

[dmelliott]

Some parts of this message have been removed. Learn more about Nabble's security policy.

 

Dear Help,

 

  Is it possible to get the actual solution points that lsode uses rather than just the regularly spaced ones generated by linspace for the input?

 

                                                                                            Thanks,

 

                                                                                                   dmelliott

_______________________________________________
Help-octave mailing list
Help-octave@...
https://www-old.cae.wisc.edu/mailman/listinfo/help-octave

[Marc Normandin]

dmelliott wrote:

> **
> *Dear Help,*
> **
> *  Is it possible to get the actual solution points that lsode uses
> rather than just the regularly spaced ones generated by linspace for the
> input?*
> **
> *                                                                      
>                     Thanks,*
> **
> *                                                                      
>                             dmelliott*

To my admittedly limited knowledge, the solution points are not
available from lsode.  You can provide lsode with an irregularly sampled
time vector, if that helps your problem any.

If you need them, the solvers in odepkg from Octave-Forge return time
stamps for each element of the solution.

--
------------------------------------------------------------------
Marc D. Normandin              http://web.ics.purdue.edu/~mdnorman
Graduate Research Assistant                     mnormand@...
Indiana University School of Medicine           317-278-9841 (tel)
Department of Radiology, Division of Research   317-274-1067 (fax)
------------------------------------------------------------------
_______________________________________________
Help-octave mailing list
Help-octave@...
https://www-old.cae.wisc.edu/mailman/listinfo/help-octave

[John W. Eaton]

On 18-Sep-2008, Marc Normandin wrote:

| dmelliott wrote:
| > **
| > *Dear Help,*
| > **
| > *  Is it possible to get the actual solution points that lsode uses
| > rather than just the regularly spaced ones generated by linspace for the
| > input?*
| > **
| > *                                                                      
| >                     Thanks,*
| > **
| > *                                                                      
| >                             dmelliott*
|
| To my admittedly limited knowledge, the solution points are not
| available from lsode.  You can provide lsode with an irregularly sampled
| time vector, if that helps your problem any.
|
| If you need them, the solvers in odepkg from Octave-Forge return time
| stamps for each element of the solution.

Or, hack Octave's lsode function to return the data you need.  Then
when you've done that, consider contributing your changes back to the
Octave project.

Thanks,

jwe
_______________________________________________
Help-octave mailing list
Help-octave@...
https://www-old.cae.wisc.edu/mailman/listinfo/help-octave

[iason]

dmelliott wrote:

Dear Help,

  Is it possible to get the actual solution points that lsode uses rather than just the regularly spaced ones generated by linspace for the input?

                                                                                            Thanks,

                                                                                                   dmelliott
_______________________________________________
Help-octave mailing list
[e-mail]
https://www-old.cae.wisc.edu/mailman/listinfo/help-octave

dmelliott,

have you achieved any success in retrieving the solution points (the independent variable nodes) from the LSODE function of OCTAVE? don't you know, is it possible to do this using the original LSODE (or similar) function from the ODEPACK library (where it comes from)?

[iason ()]

Some time ago I found out that in theory it is possible solve this task by using scipy.integrate.ode (from SciPy, not from Octave). But I could not find a way to get a solution from this function in admissible time. The function scipy.integrate.ode uses VODE instead of LSODE and allows to solve the ODE problem step-by-step.

But when I tried to use the code from NumPy samples I found out that it is too slow for practical uses. I posted a question on this matter here

http://old.nabble.com/A-discrepancy-between-NumPy-documentation-and-recomendations-for-beginers-td26765626.html

and wish to find some useful reply. By the way, don't you know, is it possible to port ode15s function (Matlab) to NumPy or Octave or SciLab? The manual and some articles I found say that ode15s function uses BDF algorithm (on user demand) and similar NDF algorithm (by default). Is there any technical or legal problem that prevents one to create an OpenSource equivalent of this very helpful function?

Marc Normandin wrote:

dmelliott wrote:
> *  Is it possible to get the actual solution points that lsode uses
> rather than just the regularly spaced ones generated by linspace for the
> input?*