Octave for windows goes mad after long simulations

Give some feedback on openEMS.

Moderators: thorsten, sebastian

Post Reply
Hale_812
Posts: 171
Joined: Fri 13 May 2016, 02:54

Octave for windows goes mad after long simulations

Post by Hale_812 » Wed 15 Jun 2016, 11:03

I have a simulation like this one.

FDTD simulation size: 85x632x23 --> 1.23556e+006 FDTD cells
FDTD timestep is: 4.25416e-014 s; Nyquist rate: 652 timesteps @1.80264e+010 Hz
Excitation signal length is: 9620 timesteps (4.0925e-010s)
Max. number of timesteps: 1000000000 ( --> 103950 * Excitation signal length)

The structure has a bandgap and when I do not know exact resonance frequency, it can finish without converging. Normally without an error, but with wrong S-params after many hours.

The strange thing is that it kills Octave.
with linspace(f_start,f_stop,801), I can not find minimums with min().

After investigation, I found that octave starts throwing "error: subscript indices must be either positive integers less than 2^31 or logicals" error even when calling min(2:8) !

max(2:8) works well, however.

Whats wrong? Is it OpenEMS influencing Octave engine, or some strange Octave behavior I am not familiar with?

thorsten
Posts: 1393
Joined: Mon 27 Jun 2011, 12:26

Re: Octave for windows goes mad after long simulations

Post by thorsten » Wed 15 Jun 2016, 22:18

Did you maybe use "min" as a variable? Because in this case you overwrite the internal function. That would explain your error message...

Hale_812
Posts: 171
Joined: Fri 13 May 2016, 02:54

Re: Octave for windows goes mad after long simulations

Post by Hale_812 » Thu 16 Jun 2016, 05:22

ah, sorry. case is closed. I have overloaded the min() in the script and other dependent functions went mad as well. :oops: I did not know, core functions can be overloaded that easily; now I know.

Post Reply