Tom O'Haver

Dan,

As you requested, I have created variants of findpeaks that compute the peak start and end positions. See

http://terpconnect.umd.edu/~toh/spectrum/PeakFindingandMeasurement.htm#PeakStartAndEnd
I'll include these in the next update of fileexchange entry.
Tom

Version 2.8 has an improved frequency Spectrum function that remains active when changing the smoothing, differentiation, and other functions, allowing you to see the effect on the frequency spectrum immediately. Press Shift-S again to return to the normal mode.

Version 9.2, uploaded today, has bug fixes; fixed-position Gaussians (shape 16) and fixed-position Lorentzians (shape 17); adds a peak shape selection menu (activated by the '-' key).

Hi Tom-

This is a great tool.

I have a question about the constant '0.600516' that appears in the function 'gaussian'. In the comment for that function, you write "half-width=wid". If you define your gaussian as:

exp(-x^2/(2*sigma)^2)

then the half width = sqrt(2log(2))*sigma = 0.776*sigma. If you define your gaussian as:

exp(-x^2/sigma^2)

then the half width = sqrt(log(2))*sigma = 0.549*sigma.

It appears that you chose the latter definition but with "wid=full width at half max". That way wid = 2*sqrt(log(2))*sigma or alternatively sigma = 0.600516*wid.

In this case, the comment should read "FWHM=wid".

Thanks
Spencer

Thanks, Avigdor,
The #9 exppulse function itself (starting on line 1152 in version 3.6) is a unit-height function, as are all the shape functions in peakfit.m. The amplitude (peak height) is supplied by the fitting function ("fitexppulse") beginning on line 1140; it's called PEAKHEIGHT. That's because the amplitude is a linear variable that is calculated by linear regression (in line 1148) rather than by iteration using fminsearch.