| Re: Doing a DFT in IDL [message #79772] |
Tue, 03 April 2012 02:23 |
Helder Marchetto
Messages: 520
Registered: November 2011
|
Senior Member |
|
|
On Monday, April 2, 2012 11:24:15 PM UTC+2, Kenneth P. Bowman wrote:
>> Dear all,
>> I've been looking for this around, but couldn't yet find anything.
>> I would like to made a (discrete) Fourier transform by directly typing in the
>> code. The target is not performance, but eventually modifying the source
>> (integrated) function from [f(x)exp(-j2piux/N)] to something different.
>> Has anybody written such a code in IDL that can be modified? Or can anybody
>> give me a working link to such a code?
>>
>> Many thanks and sorry for wasting your time.
>> Cheers, Helder
>
> A DFT is just a set of dot products. (And an FFT is an efficient way to
> carry out the DFT calculations.)
>
> Compute the basis functions that you want at the tabulated points,
> then multiply by the function you are transforming and sum.
> You can write it as a matrix-vector multiplication.
>
> Ken Bowman
Dear Ken,
thanks for your reply. I think I will have to write it, but to avoid beginners errors, I wanted to modify an available code. That's all.
Cheers, Helder
|
|
|
|
| Re: Doing a DFT in IDL [message #79773 is a reply to message #79772] |
Tue, 03 April 2012 02:21  |
Helder Marchetto
Messages: 520
Registered: November 2011
|
Senior Member |
|
|
On Monday, April 2, 2012 4:19:45 PM UTC+2, Russell wrote:
> On Apr 2, 8:31 am, Helder <hel...@marchetto.de> wrote:
>> Dear all,
>> I've been looking for this around, but couldn't yet find anything.
>> I would like to made a (discrete) Fourier transform by directly typing in the code. The target is not performance, but eventually modifying the source (integrated) function from [f(x)exp(-j2piux/N)] to something different.
>> Has anybody written such a code in IDL that can be modified? Or can anybody give me a working link to such a code?
>>
>> Many thanks and sorry for wasting your time.
>> Cheers, Helder
>
> What do you need to do that fft.pro or convol.pro won't do --- or even
> convolve.pro (with the Astronomy Library)?
>
> -Russell
Hi Russel,
I couldn't find any "fft.pro", but I think that I will go with writing my own procedure if none other is available. I was just surprised that such a demo or tutorial code is not on the web.
Cheers, Helder
|
|
|
|
| Re: Doing a DFT in IDL [message #79776 is a reply to message #79773] |
Mon, 02 April 2012 14:24 |
Kenneth P. Bowman
Messages: 585
Registered: May 2000
|
Senior Member |
|
|
In article < 22647522.376.1333369882343.JavaMail.geo-discussion-forums@yn gr3 >,
Helder <helder@marchetto.de> wrote:
> Dear all,
> I've been looking for this around, but couldn't yet find anything.
> I would like to made a (discrete) Fourier transform by directly typing in the
> code. The target is not performance, but eventually modifying the source
> (integrated) function from [f(x)exp(-j2piux/N)] to something different.
> Has anybody written such a code in IDL that can be modified? Or can anybody
> give me a working link to such a code?
>
> Many thanks and sorry for wasting your time.
> Cheers, Helder
A DFT is just a set of dot products. (And an FFT is an efficient way to
carry out the DFT calculations.)
Compute the basis functions that you want at the tabulated points,
then multiply by the function you are transforming and sum.
You can write it as a matrix-vector multiplication.
Ken Bowman
|
|
|
|
| Re: Doing a DFT in IDL [message #79778 is a reply to message #79776] |
Mon, 02 April 2012 07:19 |
Russell[1]
Messages: 101
Registered: August 2011
|
Senior Member |
|
|
On Apr 2, 8:31 am, Helder <hel...@marchetto.de> wrote:
> Dear all,
> I've been looking for this around, but couldn't yet find anything.
> I would like to made a (discrete) Fourier transform by directly typing in the code. The target is not performance, but eventually modifying the source (integrated) function from [f(x)exp(-j2piux/N)] to something different.
> Has anybody written such a code in IDL that can be modified? Or can anybody give me a working link to such a code?
>
> Many thanks and sorry for wasting your time.
> Cheers, Helder
What do you need to do that fft.pro or convol.pro won't do --- or even
convolve.pro (with the Astronomy Library)?
-Russell
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
