Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.graphics.apps.gnuplot > #3315

Re: Gaussian fit - Singular matrix in Invert_RtR

Show all headers | View raw

Am 27.04.2016 um 23:20 schrieb Andrew Finlay:
> On Wednesday, April 27, 2016 at 1:29:37 PM UTC-7, Hans-Bernhard
> Bröker wrote:

>> So let's see why this particular fit failed.  Well, for starters
>> the data is quite clearly not Gaussian at all.  It's highly
>> asymmetrical, and the tails are way to wide for compared to the
>> FWHM.
>>
>> Trying to fit a Gaussian to that data is an exercise in futility.
>
> I would argue that the peak is Gaussian to first order approximation.

It's not.  And if it were, the first order of business would have been 
to limit the fit to the region of the peak.  The dataset has 40 bins, 
but at most 6 of them could possibly be said to look Gaussian.  Fitting 
3 parameters to those 6 inputs might appear to work, and gnuplot will 
eventually do it.  It still doesn't make a useful fit, though.

Oh, and two more things: The starting value of StdDev appears about a 
factor of 2 off.

And for a StdDev around 0.005, a bin width of 0.004 can't really work, 
either.

Back to comp.graphics.apps.gnuplot | Previous | Next — Previous in thread | Next in thread | Find similar

Thread

Gaussian fit - Singular matrix in Invert_RtR Andrew Finlay <andrewlfinlay@gmail.com> - 2016-04-27 10:38 -0700
  Re: Gaussian fit - Singular matrix in Invert_RtR Hans-Bernhard Bröker <HBBroeker@t-online.de> - 2016-04-27 22:29 +0200
    Re: Gaussian fit - Singular matrix in Invert_RtR Andrew Finlay <andrewlfinlay@gmail.com> - 2016-04-27 14:20 -0700
      Re: Gaussian fit - Singular matrix in Invert_RtR Hans-Bernhard Bröker <HBBroeker@t-online.de> - 2016-04-28 00:51 +0200
  Re: Gaussian fit - Singular matrix in Invert_RtR boettcher@physik.uni-kiel.de (Stephan Böttcher) - 2016-04-28 07:30 +0200

csiph-web