Re: [ferret_users] Spatial Fourier Analysis

["Karl M. Smith" <karl.smith@xxxxxxxx>]

Hi Dave,

Currently the FFT functions all perform the transformation on a time 
axis.  One could write an external function to do this on other axes, or 
one could regrid to make the desired axis a (fake) time axis.

However, I am writing to see another option might be of interest: 
PyFerret (still not officially released, but working quite well) which 
is Ferret built as a Python module.  If you are familiar with Python, 
using the numpy.fft module will allow you to do your FFT analysis in 
Python in a multitude of ways.  I can envision writing a Python based 
external function that gets the data from Ferret, uses the numpy.fft 
package to (a) transform the passed data to the frequency domain along 
whichever axis you desire, (b) smooth or otherwise modify the 
frequencies, and (c) transform the data back, and then return this 
modified data array.  (There is also the option of using PyFerret's 
getdata and putdata methods to get data arrays from Ferret and put 
whatever data arrays you want back into Ferret.)

Let me know if you are interested in giving PyFerret a try.

Karl


David Ogier wrote:
> Hi Ferreters...
>
> I've been using FFTA to do fourier analysis on some data, focusing on 
> the time aspect for the frequency, but I'd like to do a spatial 
> fourier analysis where frequencies are mapped in the x or y rather 
> then with time?
> Is this possible? can ffta be used for this?
>
> Many thanks
>
> Dave

--
Karl M. Smith, Ph.D.
Research Scientist and Software Developer
UW/JISAO and NOAA/PMEL/TMAP
karl.smith@xxxxxxxx
(206) 526-4806

"The contents of this message are mine personally and do
not necessarily reflect any position of the Government
or the National Oceanic and Atmospheric Administration."