[nSLUG] Re: Possibly ignorant quest for simple FFT app

Mike Spencer mspencer at tallships.ca
Sat Jul 13 03:39:38 ADT 2013



Jack W. wrote:

> First, there is a clear distinction between a Fourier *series*
> (right direction) and a Fourier *transform* (blind alley).

Ah.  Very good.  I was inferring that some embodiment of FT would be
give me  the coefficients of the terms of an F series.  I think you're
telling me I'm wrong to think that.

> Digression: I am sure you are familiar with so-called "linear
> regression", where a straight line is fitted to a set of data by the
> method of "least squares".

Yes, or rather, I once was. I know which book to look in to refresh my
memory.  I *think* I can find the book. :-)

> You want to do a different kind of curve fitting.

Well, sort of.

I want to look for periodicity in an array of values that may or may
not have any.  In the case of the mouse, I want to see if there's a
*regular* periodic component in his episodic sprints or within the
sprint episodes.

This is somewhat similar to looking for periodicity in a list of
putative random numbers. I once rigged up a web cam looking at a
plasma globe and tried to extract random numbers from the images.
Didn't work worth a pinch of fuvg (at least for a first
approximation).  The behavior of the plasma was evidently stochastic
but the imaging process defeated the intent.  But my attention was
directed to the notion of looking for periodicity in apparent
disorder.

Your suggestion that I'm looking for the wrong thing leads me to
grovel further.  Maybe a "power spectrum" tool is what I really want.
Still groveling.

> GNU Plot looked especially promising.

I've used gnuplot quite a bit. It's great for visualizing data but
mostly doesn't process it unless the user can write the processing as
gnuplot algebraic expressions. For data images that are too squinched
up when fitted to the screen, gnuplot allows you to write to an image
file of arbitrary size that can then be viewed in an image viewer or
browser.


Thanks, and for the  suggestions from others, too. So far, nothing has
turned up that is simpleminded enough for me.

- Mike

-- 
Michael Spencer                  Nova Scotia, Canada       .~. 
                                                           /V\ 
mspencer at tallships.ca                                     /( )\
http://home.tallships.ca/mspencer/                        ^^-^^


More information about the nSLUG mailing list