Any Octave users? What is this code snippet actually doing?

[Bruce Dawson]

Unless matlab treats floats and ints differently than most other
languages, I'd say its a bug because zero over anything is always zero.

But, I have no clue as to what context its running in, so it may be
deliberate.

--Bruce

On 04/06/2013 02:03 PM, Bruce Labitt wrote:
> I'm trying to understand a chunk of 3D FDTD code that I downloaded
> from the publisher of the book, "Computational Electrodynamics".  It
> is in matlab, so I tried to run it in Octave.  Something is not quite
> right, so it doesn't execute.  That doesn't matter to me right now. 
> The piece of code is:
>
> //start snippet
> %   x-varying material properties
> delbc=upml*delta;
> sigmam=-log(rmax)*(orderbc+1.0)/(2.0*eta*delbc); 
> sigfactor=sigmam/(delta*(delbc^orderbc)*(orderbc+1.0));
> kmax=1;
> kfactor=(kmax-1.0)/delta/(orderbc+1.0)/delbc^orderbc;
> //end snippet
>
> It would seem to me that kfactor is always zero, since kmax is set to
> 1 in the line above.  That probably is not the intent, I think. 
> kfactor does not seem to be updated later in the code.
>
> Is this just a bug?  Or what does the last line do?  The author of the
> code is a long graduated PhD student from UWisc.  Just contact UW? 
>
> URL of actual code: 
> http://www.artechhouse.com/static/downloads/fdtd3D_UPML.zip
>
> It doesn't matter if it executes in octave.  (It would be nice, though
> - it currently runs, plots a little, then seg faults) 
>
> octave:1> fdtd3D_UPML
> error: base_graphics_object::get_properties: invalid graphics object
> panic: Segmentation fault -- stopping myself...
> attempting to save variables to `octave-core'...
> save to `octave-core' complete
> Segmentation fault
>
> I tested the code in Matlab, and it executes.  Can't really say if the
> code is doing everything right, but it runs and vaguely does things
> sort of ok. The simulation is kind of coarse, but one can see an
> electric field starting to radiate from a dipole.
>
> I'm porting it to python/scipy and maybe later to pycuda, once I get
> it working.  Then maybe I can do some cool stuff with it.
>
> Thanks in Advance
> -Bruce
>
>
>
>
> _______________________________________________
> gnhlug-discuss mailing list
> gnhlug-discuss at mail.gnhlug.org
> http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.gnhlug.org/mailman/private/gnhlug-discuss/attachments/20130406/2ba6dae7/attachment.html