Thread Subject: functional basis representation

hi all,
am having about 450000 samples in say 'Y' matrix. how could i represent 'Y' in wavelet base,fourier base.
please someone help me.

thanks,
matsya

"matsya mopad" <matsya89@gmail.com> wrote in message <hi949e$g4i$1@fred.mathworks.com>...
> hi all,
> am having about 450000 samples in say 'Y' matrix. how could i represent 'Y' in wavelet base,fourier base.
===================

For a Fourier representation, see the fft() help.

For wavelet representation, see the wavelet toolbox user guide

http://www.mathworks.com/access/helpdesk/help/toolbox/wavelet/wavelet_product_page.html


If you do not have the wavelet toolbox, but you do have a way of generating wavelet basis function samples, you could also use my interpMatrix tool to build a wavelet basis matrix:


http://www.mathworks.com/matlabcentral/fileexchange/26292-regular-control-point-interpolation-matrix-with-boundary-conditions

You could then compute the transform using matrix inversion methods.
However, you would have to construct such a matrix for each scale that you wished to process. I can't say anything about the efficiency of this approach relative to an actual DWT. At coarse scales, it would probably be more efficient for you to use fft() to implement your own filter-and-decimate routine.

In addition, there are quite a few wavelet analysis tools on the file exchange:

http://www.mathworks.com/matlabcentral/fileexchange/?term=wavelet

Same thing for Fourier analysis.

Hi Matsya,

You could use the Atomizer to find which bases are closed to your signal.
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.52.2746

Regards,
Nicolas

"Nicolas " <cusseani@ensieta.fr> wrote in message <hic7bf$cub$1@fred.mathworks.com>...
> Hi Matsya,
>
> You could use the Atomizer to find which bases are closed to your signal.
> http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.52.2746
>
> Regards,
> Nicolas

thanks a lot matt j and nicolas.ur suggested links were very helpful.