trivial deconvolution -- possible without fourier?

Hello,

I've performed a trivial convolution on a set of one-dimensional data,
with a gaussian impluse function. I can then deconvolve the resultant
set of data with the same gaussian to get back, 100% precisely, the
original data. I'm using a simple piece of software called SPECTRUM
that does this by multiplying or dividing the fourier analysis of the
two functions.

I need to implement the deconvolution side in code of my own, but would
like to not have to bother with FFT if I don't have to. Is there any
simpler method of doing such a trivial deconvolution just using
standard linear algebra or something else? If someone could point me in
the right direction, I'd really appreciate it.

Thanks
Michael

.

Relevant Pages

  • Re: trivial deconvolution -- possible without fourier?
    ... I've performed a trivial convolution on a set of one-dimensional data, ... with a gaussian impluse function. ... I need to implement the deconvolution side in code of my own, ... My guess is that if both convolutes are one sided (i.e. only exist for n>=0 so that the fist term of the convolution only has one term in it) then you can solve it - it will be a bit like taking the inverse of a lower triangular matrix. ...
    (sci.math)
  • trivial deconvolution -- possible without fourier?
    ... I've performed a trivial convolution on a set of one-dimensional data, ... with a gaussian impluse function. ... I can then deconvolve the resultant ... I need to implement the deconvolution side in code of my own, ...
    (sci.math)
  • Re: trivial deconvolution -- possible without fourier?
    ... I've performed a trivial convolution on a set of one-dimensional data, ... with a gaussian impluse function. ... I need to implement the deconvolution side in code of my own, ...
    (sci.math.num-analysis)
Loading