[clean-list] FFT with Clean

[Siegfried Gonzi]

Antonio Eduardo Costa Pereira wrote:

Dear Antonio Eduardo,

> >And for that simulation I need the FFT (with a resolution up to
> >1024x1024 array points).
>
> That will be easy. Carlos used arrays as big as 65535, to produce
>  results in less than one second.
>

Are you shure? When I type for example 16328 instead of 4095 I wait and
wait and wait 12sec!

Yorick for example tooks for a 1024x1024 array 50sec for the FFT on my
old Mac.

> In my youth, I worked at Cornell, with people who used to make
> observations from Arecibo's Radiotelescope (Carl Sagan,
> Donald Farley, Mike Kelley, Bela Fejer were my bosses :).

I am only 26 years old, but I know that in 1976 Forth was the official
programming language for the International-Astronomical-Union (IAU). I
know an Italian radio-astronomer who really programmend in Forth his
radio-telescope. He was happy to know that I am  try to learn Forth now,
but the next time I will give him bad news...

> You are right.
> I would say that, if you want to control the telescope, Forth
> is not that bad. For anything else, you should go for
> another tool.
>

I was (first) impressed by Forth, because there are a physicist who
claims for Forth. He also often writes articles about Forth (the last
was in the peer-reviewed magazine: "Computing in Science and
Engineering", Sep./Oct. 2000, p. 6ff, from J.Noble.").

And I often red that some people want to get people from Fortran90 to
Forth. Oh, Oh, Forth is slow and slow. For example a tuned (with
subarray tricks) matrix-matrix-multplication takes 3 to 4 times longer
than the multiplication with Clean.

>
> By the way, there is a thing that may interest you.
> Ferraz de Mello, a well known astronomer, invented a
> variation of the Fourier Transform that is largely
> used by people who work with Variable Stars. You can
> access the site of the American Association of Variable
> Stars Observers (AAVSO) (http://www.aavso.org/) for details.

Thank you.

> A last thing. You can use Clean even to control your
> telescope (I suppose that you are an astronomer).
> All you need to do is to write a module in
> C to take care of low level control and data acquisition.
> If you want, I can send you the code you need to
> access devices from Linux.

We have a programmer, who programms the telescope-control in C. The
telescope (CCD camera with interference filters. eg. CaIIK-line for the
Sun's chromosphere). The telescope should than be incorporated in an
existing infrastructure, that means I do not have the possibility to
make changes.
And it would be a great problem to persuade my boss, they all believe
here IDL and C are the only languages around. I hate IDL and for object
oriented programming (like C++) I am to stupid.

Maybe when I start (next year) with my PhD I have more freedom...

But the problem often is that there are big IDL libraries around, for
example we have to flat-fielding our images from the Sun after the
Kun-Lin-method. And such things often are available in IDL. We use for
flat-fielding the same IDL library as at Big-Bear solar observatory.

> Sorry, I do not work with MacIntosh.

No problem. I also only can work with my Mac at home. Here at the
university we have Suns (though old ones) and NT.


Regards,
S. Gonzi

>