1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
|
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html>
<!-- Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013 The GSL Team.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.3 or
any later version published by the Free Software Foundation; with no
Invariant Sections and no cover texts. A copy of the license is
included in the section entitled "GNU Free Documentation License". -->
<!-- Created by GNU Texinfo 5.1, http://www.gnu.org/software/texinfo/ -->
<head>
<title>GNU Scientific Library – Reference Manual: Fitting Overview</title>
<meta name="description" content="GNU Scientific Library – Reference Manual: Fitting Overview">
<meta name="keywords" content="GNU Scientific Library – Reference Manual: Fitting Overview">
<meta name="resource-type" content="document">
<meta name="distribution" content="global">
<meta name="Generator" content="makeinfo">
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<link href="index.html#Top" rel="start" title="Top">
<link href="Function-Index.html#Function-Index" rel="index" title="Function Index">
<link href="Least_002dSquares-Fitting.html#Least_002dSquares-Fitting" rel="up" title="Least-Squares Fitting">
<link href="Linear-regression.html#Linear-regression" rel="next" title="Linear regression">
<link href="Least_002dSquares-Fitting.html#Least_002dSquares-Fitting" rel="previous" title="Least-Squares Fitting">
<style type="text/css">
<!--
a.summary-letter {text-decoration: none}
blockquote.smallquotation {font-size: smaller}
div.display {margin-left: 3.2em}
div.example {margin-left: 3.2em}
div.indentedblock {margin-left: 3.2em}
div.lisp {margin-left: 3.2em}
div.smalldisplay {margin-left: 3.2em}
div.smallexample {margin-left: 3.2em}
div.smallindentedblock {margin-left: 3.2em; font-size: smaller}
div.smalllisp {margin-left: 3.2em}
kbd {font-style:oblique}
pre.display {font-family: inherit}
pre.format {font-family: inherit}
pre.menu-comment {font-family: serif}
pre.menu-preformatted {font-family: serif}
pre.smalldisplay {font-family: inherit; font-size: smaller}
pre.smallexample {font-size: smaller}
pre.smallformat {font-family: inherit; font-size: smaller}
pre.smalllisp {font-size: smaller}
span.nocodebreak {white-space:nowrap}
span.nolinebreak {white-space:nowrap}
span.roman {font-family:serif; font-weight:normal}
span.sansserif {font-family:sans-serif; font-weight:normal}
ul.no-bullet {list-style: none}
-->
</style>
</head>
<body lang="en" bgcolor="#FFFFFF" text="#000000" link="#0000FF" vlink="#800080" alink="#FF0000">
<a name="Fitting-Overview"></a>
<div class="header">
<p>
Next: <a href="Linear-regression.html#Linear-regression" accesskey="n" rel="next">Linear regression</a>, Up: <a href="Least_002dSquares-Fitting.html#Least_002dSquares-Fitting" accesskey="u" rel="up">Least-Squares Fitting</a> [<a href="Function-Index.html#Function-Index" title="Index" rel="index">Index</a>]</p>
</div>
<hr>
<a name="Overview-4"></a>
<h3 class="section">37.1 Overview</h3>
<p>Least-squares fits are found by minimizing <em>\chi^2</em>
(chi-squared), the weighted sum of squared residuals over <em>n</em>
experimental datapoints <em>(x_i, y_i)</em> for the model <em>Y(c,x)</em>,
The <em>p</em> parameters of the model are <em>c = {c_0, c_1, …}</em>. The
weight factors <em>w_i</em> are given by <em>w_i = 1/\sigma_i^2</em>,
where <em>\sigma_i</em> is the experimental error on the data-point
<em>y_i</em>. The errors are assumed to be
Gaussian and uncorrelated.
For unweighted data the chi-squared sum is computed without any weight factors.
</p>
<p>The fitting routines return the best-fit parameters <em>c</em> and their
<em>p \times p</em> covariance matrix. The covariance matrix measures the
statistical errors on the best-fit parameters resulting from the
errors on the data, <em>\sigma_i</em>, and is defined
<a name="index-covariance-matrix_002c-linear-fits"></a>
as <em>C_{ab} = <\delta c_a \delta c_b></em> where <em>< ></em> denotes an average over the Gaussian error distributions of the underlying datapoints.
</p>
<p>The covariance matrix is calculated by error propagation from the data
errors <em>\sigma_i</em>. The change in a fitted parameter <em>\delta
c_a</em> caused by a small change in the data <em>\delta y_i</em> is given
by
allowing the covariance matrix to be written in terms of the errors on the data,
For uncorrelated data the fluctuations of the underlying datapoints satisfy
<em><\delta y_i \delta y_j> = \sigma_i^2 \delta_{ij}</em>, giving a
corresponding parameter covariance matrix of
When computing the covariance matrix for unweighted data, i.e. data with unknown errors,
the weight factors <em>w_i</em> in this sum are replaced by the single estimate <em>w =
1/\sigma^2</em>, where <em>\sigma^2</em> is the computed variance of the
residuals about the best-fit model, <em>\sigma^2 = \sum (y_i - Y(c,x_i))^2 / (n-p)</em>.
This is referred to as the <em>variance-covariance matrix</em>.
<a name="index-variance_002dcovariance-matrix_002c-linear-fits"></a>
</p>
<p>The standard deviations of the best-fit parameters are given by the
square root of the corresponding diagonal elements of
the covariance matrix, <em>\sigma_{c_a} = \sqrt{C_{aa}}</em>.
The correlation coefficient of the fit parameters <em>c_a</em> and <em>c_b</em>
is given by <em>\rho_{ab} = C_{ab} / \sqrt{C_{aa} C_{bb}}</em>.
</p>
<hr>
<div class="header">
<p>
Next: <a href="Linear-regression.html#Linear-regression" accesskey="n" rel="next">Linear regression</a>, Up: <a href="Least_002dSquares-Fitting.html#Least_002dSquares-Fitting" accesskey="u" rel="up">Least-Squares Fitting</a> [<a href="Function-Index.html#Function-Index" title="Index" rel="index">Index</a>]</p>
</div>
</body>
</html>
|
