Re: 20 GB console.log documents generated

[Dmitri A. Sergatskov]

On 10/30/07, James Sherman Jr. <address@hidden> wrote:
> On 10/29/07, cal <address@hidden> wrote:
> >
...
> Like Jordi has already said, is there any particular reason that you
> need to plot so many values?  Working with those many points is fine
> in octave, but because of how octave and gnuplot interface, its not a
> good idea to plot so many.

I am not Cal, but I do not understand since when 20000 points becomes
"too many."
I am routinely plot 10 times as much. I would like to have a rough idea what
my data look like before start messing with it. BTW on my computer the
Cal's code returns:


octave:1> tic
octave:2>
octave:2> T = 0 : 0.0001 : 2 - .0001;
octave:3> Y = 2.3*sin(2*pi*10*T);
octave:4>
octave:4> toc
Elapsed time is 0.129105 seconds.
octave:5>
octave:5> plot(T,Y)
octave:6>
octave:6> toc
Elapsed time is 1.222946 seconds.
octave:7> whos Y

*** local user variables:

  Prot Name        Size                     Bytes  Class
  ==== ====        ====                     =====  =====
   rwd Y           1x20000                 160000  double

Total is 20000 elements using 160000 bytes

octave:8>

Of course, after the last "toc" gnuplot takes some (very short) time
to plot the data octave handed to it. This is with octave-2.9.9 on
Fedora 7, so octave still writes the temp data file (0.5M).

In any case, if PPC version chocks on this, there is something
really, really wrong with that setup...

>
> James

Sincerely,

Dmitri.
--