Path: csiph.com!fu-berlin.de!uni-berlin.de!news.dfncis.de!not-for-mail
From: =?UTF-8?Q?Hans-Bernhard_Br=c3=b6ker?= <HBBroeker@t-online.de>
Newsgroups: comp.graphics.apps.gnuplot
Subject: Re: Gaussian fit - Singular matrix in Invert_RtR
Date: Wed, 27 Apr 2016 22:29:27 +0200
Lines: 15
Message-ID: <doclpfF62dsU1@mid.dfncis.de>
References: <f2adc325-2828-4a1a-9bee-47e5afb8691f@googlegroups.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
X-Trace: news.dfncis.de X7dxhQz/RmRbleiktWsHPgTuEfyDtQ/pfRZyAFfEZtJbvgv8OQkJlG7jHw
Cancel-Lock: sha1:ipUZnXFg4wm7YEoe1KqLuq53w2M=
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:45.0) Gecko/20100101 Thunderbird/45.0
In-Reply-To: <f2adc325-2828-4a1a-9bee-47e5afb8691f@googlegroups.com>
Xref: csiph.com comp.graphics.apps.gnuplot:3313

Am 27.04.2016 um 19:38 schrieb Andrew Finlay:

> I'm trying to write code to automatically fit binned simulation data

Bad idea.  'fit' is not a tool to be used blindly, or automatically.

Non-linear least-squares fitting needs guidance and consideration of the 
condition of the input, or you'll run into numerical instabilities much 
of the time.

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.