<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=US-ASCII">
<title>Modified Bessel Functions of the First and Second Kinds</title>
<link rel="stylesheet" href="../../../../../../../../doc/src/boostbook.css" type="text/css">
<meta name="generator" content="DocBook XSL Stylesheets V1.74.0">
<link rel="home" href="../../../index.html" title="Math Toolkit">
<link rel="up" href="../bessel.html" title="Bessel Functions">
<link rel="prev" href="bessel.html" title="Bessel Functions of the First and Second Kinds">
<link rel="next" href="sph_bessel.html" title="Spherical Bessel Functions of the First and Second Kinds">
</head>
<body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF">
<table cellpadding="2" width="100%"><tr>
<td valign="top"><img alt="Boost C++ Libraries" width="277" height="86" src="../../../../../../../../boost.png"></td>
<td align="center"><a href="../../../../../../../../index.html">Home</a></td>
<td align="center"><a href="../../../../../../../../libs/libraries.htm">Libraries</a></td>
<td align="center"><a href="http://www.boost.org/users/people.html">People</a></td>
<td align="center"><a href="http://www.boost.org/users/faq.html">FAQ</a></td>
<td align="center"><a href="../../../../../../../../more/index.htm">More</a></td>
</tr></table>
<hr>
<div class="spirit-nav">
<a accesskey="p" href="bessel.html"><img src="../../../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../bessel.html"><img src="../../../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../../index.html"><img src="../../../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="sph_bessel.html"><img src="../../../../../../../../doc/src/images/next.png" alt="Next"></a>
</div>
<div class="section" lang="en">
<div class="titlepage"><div><div><h4 class="title">
<a name="math_toolkit.special.bessel.mbessel"></a><a class="link" href="mbessel.html" title="Modified Bessel Functions of the First and Second Kinds"> Modified Bessel
Functions of the First and Second Kinds</a>
</h4></div></div></div>
<a name="math_toolkit.special.bessel.mbessel.synopsis"></a><h5>
<a name="id1130984"></a>
<a class="link" href="mbessel.html#math_toolkit.special.bessel.mbessel.synopsis">Synopsis</a>
</h5>
<pre class="programlisting"><span class="keyword">template</span> <span class="special"><</span><span class="keyword">class</span> <span class="identifier">T1</span><span class="special">,</span> <span class="keyword">class</span> <span class="identifier">T2</span><span class="special">></span>
<a class="link" href="../../main_overview/result_type.html" title="Calculation of the Type of the Result"><span class="emphasis"><em>calculated-result-type</em></span></a> <span class="identifier">cyl_bessel_i</span><span class="special">(</span><span class="identifier">T1</span> <span class="identifier">v</span><span class="special">,</span> <span class="identifier">T2</span> <span class="identifier">x</span><span class="special">);</span>
<span class="keyword">template</span> <span class="special"><</span><span class="keyword">class</span> <span class="identifier">T1</span><span class="special">,</span> <span class="keyword">class</span> <span class="identifier">T2</span><span class="special">,</span> <span class="keyword">class</span> <a class="link" href="../../policy.html" title="Policies">Policy</a><span class="special">></span>
<a class="link" href="../../main_overview/result_type.html" title="Calculation of the Type of the Result"><span class="emphasis"><em>calculated-result-type</em></span></a> <span class="identifier">cyl_bessel_i</span><span class="special">(</span><span class="identifier">T1</span> <span class="identifier">v</span><span class="special">,</span> <span class="identifier">T2</span> <span class="identifier">x</span><span class="special">,</span> <span class="keyword">const</span> <a class="link" href="../../policy.html" title="Policies">Policy</a><span class="special">&);</span>
<span class="keyword">template</span> <span class="special"><</span><span class="keyword">class</span> <span class="identifier">T1</span><span class="special">,</span> <span class="keyword">class</span> <span class="identifier">T2</span><span class="special">></span>
<a class="link" href="../../main_overview/result_type.html" title="Calculation of the Type of the Result"><span class="emphasis"><em>calculated-result-type</em></span></a> <span class="identifier">cyl_bessel_k</span><span class="special">(</span><span class="identifier">T1</span> <span class="identifier">v</span><span class="special">,</span> <span class="identifier">T2</span> <span class="identifier">x</span><span class="special">);</span>
<span class="keyword">template</span> <span class="special"><</span><span class="keyword">class</span> <span class="identifier">T1</span><span class="special">,</span> <span class="keyword">class</span> <span class="identifier">T2</span><span class="special">,</span> <span class="keyword">class</span> <a class="link" href="../../policy.html" title="Policies">Policy</a><span class="special">></span>
<a class="link" href="../../main_overview/result_type.html" title="Calculation of the Type of the Result"><span class="emphasis"><em>calculated-result-type</em></span></a> <span class="identifier">cyl_bessel_k</span><span class="special">(</span><span class="identifier">T1</span> <span class="identifier">v</span><span class="special">,</span> <span class="identifier">T2</span> <span class="identifier">x</span><span class="special">,</span> <span class="keyword">const</span> <a class="link" href="../../policy.html" title="Policies">Policy</a><span class="special">&);</span>
</pre>
<a name="math_toolkit.special.bessel.mbessel.description"></a><h5>
<a name="id1131351"></a>
<a class="link" href="mbessel.html#math_toolkit.special.bessel.mbessel.description">Description</a>
</h5>
<p>
The functions <a class="link" href="mbessel.html" title="Modified Bessel Functions of the First and Second Kinds">cyl_bessel_i</a>
and <a class="link" href="mbessel.html" title="Modified Bessel Functions of the First and Second Kinds">cyl_bessel_k</a>
return the result of the modified Bessel functions of the first and second
kind respectively:
</p>
<p>
cyl_bessel_i(v, x) = I<sub>v</sub>(x)
</p>
<p>
cyl_bessel_k(v, x) = K<sub>v</sub>(x)
</p>
<p>
where:
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel2.png"></span>
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel3.png"></span>
</p>
<p>
The return type of these functions is computed using the <a class="link" href="../../main_overview/result_type.html" title="Calculation of the Type of the Result"><span class="emphasis"><em>result
type calculation rules</em></span></a> when T1 and T2 are different types.
The functions are also optimised for the relatively common case that T1
is an integer.
</p>
<p>
</p>
<p>
The final <a class="link" href="../../policy.html" title="Policies">Policy</a> argument
is optional and can be used to control the behaviour of the function:
how it handles errors, what level of precision to use etc. Refer to the
<a class="link" href="../../policy.html" title="Policies">policy documentation for more details</a>.
</p>
<p>
</p>
<p>
The functions return the result of <a class="link" href="../../main_overview/error_handling.html#domain_error">domain_error</a>
whenever the result is undefined or complex. For <a class="link" href="bessel.html" title="Bessel Functions of the First and Second Kinds">cyl_bessel_j</a>
this occurs when <code class="computeroutput"><span class="identifier">x</span> <span class="special"><</span>
<span class="number">0</span></code> and v is not an integer, or when
<code class="computeroutput"><span class="identifier">x</span> <span class="special">==</span>
<span class="number">0</span></code> and <code class="computeroutput"><span class="identifier">v</span>
<span class="special">!=</span> <span class="number">0</span></code>.
For <a class="link" href="bessel.html" title="Bessel Functions of the First and Second Kinds">cyl_neumann</a>
this occurs when <code class="computeroutput"><span class="identifier">x</span> <span class="special"><=</span>
<span class="number">0</span></code>.
</p>
<p>
The following graph illustrates the exponential behaviour of I<sub>v</sub>.
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../graphs/cyl_bessel_i.png" align="middle"></span>
</p>
<p>
The following graph illustrates the exponential decay of K<sub>v</sub>.
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../graphs/cyl_bessel_k.png" align="middle"></span>
</p>
<a name="math_toolkit.special.bessel.mbessel.testing"></a><h5>
<a name="id1131623"></a>
<a class="link" href="mbessel.html#math_toolkit.special.bessel.mbessel.testing">Testing</a>
</h5>
<p>
There are two sets of test values: spot values calculated using <a href="http://functions.wolfram.com" target="_top">functions.wolfram.com</a>, and a
much larger set of tests computed using a simplified version of this implementation
(with all the special case handling removed).
</p>
<a name="math_toolkit.special.bessel.mbessel.accuracy"></a><h5>
<a name="id1131645"></a>
<a class="link" href="mbessel.html#math_toolkit.special.bessel.mbessel.accuracy">Accuracy</a>
</h5>
<p>
The following tables show how the accuracy of these functions varies on
various platforms, along with a comparison to the <a href="http://www.gnu.org/software/gsl/" target="_top">GSL-1.9</a>
library. Note that only results for the widest floating-point type on the
system are given, as narrower types have <a class="link" href="../../backgrounders/relative_error.html#zero_error">effectively
zero error</a>. All values are relative errors in units of epsilon.
</p>
<div class="table">
<a name="id1131669"></a><p class="title"><b>Table 38. Errors Rates in cyl_bessel_i</b></p>
<div class="table-contents"><table class="table" summary="Errors Rates in cyl_bessel_i">
<colgroup>
<col>
<col>
<col>
</colgroup>
<thead><tr>
<th>
<p>
Significand Size
</p>
</th>
<th>
<p>
Platform and Compiler
</p>
</th>
<th>
<p>
I<sub>v</sub>
</p>
</th>
</tr></thead>
<tbody>
<tr>
<td>
<p>
53
</p>
</td>
<td>
<p>
Win32 / Visual C++ 8.0
</p>
</td>
<td>
<p>
Peak=10 Mean=3.4 GSL Peak=6000
</p>
</td>
</tr>
<tr>
<td>
<p>
64
</p>
</td>
<td>
<p>
Red Hat Linux IA64 / G++ 3.4
</p>
</td>
<td>
<p>
Peak=11 Mean=3
</p>
</td>
</tr>
<tr>
<td>
<p>
64
</p>
</td>
<td>
<p>
SUSE Linux AMD64 / G++ 4.1
</p>
</td>
<td>
<p>
Peak=11 Mean=4
</p>
</td>
</tr>
<tr>
<td>
<p>
113
</p>
</td>
<td>
<p>
HP-UX / HP aCC 6
</p>
</td>
<td>
<p>
Peak=15 Mean=4
</p>
</td>
</tr>
</tbody>
</table></div>
</div>
<br class="table-break"><div class="table">
<a name="id1131973"></a><p class="title"><b>Table 39. Errors Rates in cyl_bessel_k</b></p>
<div class="table-contents"><table class="table" summary="Errors Rates in cyl_bessel_k">
<colgroup>
<col>
<col>
<col>
</colgroup>
<thead><tr>
<th>
<p>
Significand Size
</p>
</th>
<th>
<p>
Platform and Compiler
</p>
</th>
<th>
<p>
K<sub>v</sub>
</p>
</th>
</tr></thead>
<tbody>
<tr>
<td>
<p>
53
</p>
</td>
<td>
<p>
Win32 / Visual C++ 8.0
</p>
</td>
<td>
<p>
Peak=9 Mean=2
</p>
<p>
GSL Peak=9
</p>
</td>
</tr>
<tr>
<td>
<p>
64
</p>
</td>
<td>
<p>
Red Hat Linux IA64 / G++ 3.4
</p>
</td>
<td>
<p>
Peak=10 Mean=2
</p>
</td>
</tr>
<tr>
<td>
<p>
64
</p>
</td>
<td>
<p>
SUSE Linux AMD64 / G++ 4.1
</p>
</td>
<td>
<p>
Peak=10 Mean=2
</p>
</td>
</tr>
<tr>
<td>
<p>
113
</p>
</td>
<td>
<p>
HP-UX / HP aCC 6
</p>
</td>
<td>
<p>
Peak=12 Mean=5
</p>
</td>
</tr>
</tbody>
</table></div>
</div>
<br class="table-break"><a name="math_toolkit.special.bessel.mbessel.implementation"></a><h5>
<a name="id1132135"></a>
<a class="link" href="mbessel.html#math_toolkit.special.bessel.mbessel.implementation">Implementation</a>
</h5>
<p>
The following are handled as special cases first:
</p>
<p>
When computing I<sub>v</sub> for <span class="emphasis"><em>x < 0</em></span>, then ν must be an integer
or a domain error occurs. If ν is an integer, then the function is odd if
ν is odd and even if ν is even, and we can reflect to <span class="emphasis"><em>x > 0</em></span>.
</p>
<p>
For I<sub>v</sub> with v equal to 0, 1 or 0.5 are handled as special cases.
</p>
<p>
The 0 and 1 cases use minimax rational approximations on finite and infinite
intervals. The coefficients are from:
</p>
<div class="itemizedlist"><ul type="disc">
<li>
J.M. Blair and C.A. Edwards, <span class="emphasis"><em>Stable rational minimax approximations
to the modified Bessel functions I_0(x) and I_1(x)</em></span>, Atomic
Energy of Canada Limited Report 4928, Chalk River, 1974.
</li>
<li>
S. Moshier, <span class="emphasis"><em>Methods and Programs for Mathematical Functions</em></span>,
Ellis Horwood Ltd, Chichester, 1989.
</li>
</ul></div>
<p>
While the 0.5 case is a simple trigonometric function:
</p>
<p>
I<sub>0.5</sub>(x) = sqrt(2 / πx) * sinh(x)
</p>
<p>
For K<sub>v</sub> with <span class="emphasis"><em>v</em></span> an integer, the result is calculated
using the recurrence relation:
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel5.png"></span>
</p>
<p>
starting from K<sub>0</sub> and K<sub>1</sub> which are calculated using rational the approximations
above. These rational approximations are accurate to around 19 digits,
and are therefore only used when T has no more than 64 binary digits of
precision.
</p>
<p>
In the general case, we first normalize ν to [<code class="literal">0, [inf</code>])
with the help of the reflection formulae:
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel9.png"></span>
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel10.png"></span>
</p>
<p>
Let μ = ν - floor(ν + 1/2), then μ is the fractional part of ν such that |μ| <=
1/2 (we need this for convergence later). The idea is to calculate K<sub>μ</sub>(x)
and K<sub>μ+1</sub>(x), and use them to obtain I<sub>ν</sub>(x) and K<sub>ν</sub>(x).
</p>
<p>
The algorithm is proposed by Temme in N.M. Temme, <span class="emphasis"><em>On the numerical
evaluation of the modified bessel function of the third kind</em></span>,
Journal of Computational Physics, vol 19, 324 (1975), which needs two continued
fractions as well as the Wronskian:
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel11.png"></span>
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel12.png"></span>
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel8.png"></span>
</p>
<p>
The continued fractions are computed using the modified Lentz's method
(W.J. Lentz, <span class="emphasis"><em>Generating Bessel functions in Mie scattering calculations
using continued fractions</em></span>, Applied Optics, vol 15, 668 (1976)).
Their convergence rates depend on <span class="emphasis"><em>x</em></span>, therefore we
need different strategies for large <span class="emphasis"><em>x</em></span> and small <span class="emphasis"><em>x</em></span>.
</p>
<p>
<span class="emphasis"><em>x > v</em></span>, CF1 needs O(<span class="emphasis"><em>x</em></span>) iterations
to converge, CF2 converges rapidly.
</p>
<p>
<span class="emphasis"><em>x <= v</em></span>, CF1 converges rapidly, CF2 fails to converge
when <span class="emphasis"><em>x</em></span> <code class="literal">-></code> 0.
</p>
<p>
When <span class="emphasis"><em>x</em></span> is large (<span class="emphasis"><em>x</em></span> > 2), both
continued fractions converge (CF1 may be slow for really large <span class="emphasis"><em>x</em></span>).
K<sub>μ</sub> and K<sub>μ+1</sub>
can be calculated by
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel13.png"></span>
</p>
<p>
where
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel14.png"></span>
</p>
<p>
<span class="emphasis"><em>S</em></span> is also a series that is summed along with CF2,
see I.J. Thompson and A.R. Barnett, <span class="emphasis"><em>Modified Bessel functions
I_v and K_v of real order and complex argument to selected accuracy</em></span>,
Computer Physics Communications, vol 47, 245 (1987).
</p>
<p>
When <span class="emphasis"><em>x</em></span> is small (<span class="emphasis"><em>x</em></span> <= 2),
CF2 convergence may fail (but CF1 works very well). The solution here is
Temme's series:
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel15.png"></span>
</p>
<p>
where
</p>
<p>
<span class="inlinemediaobject"><img src="../../../../equations/mbessel16.png"></span>
</p>
<p>
f<sub>k</sub> and h<sub>k</sub>
are also computed by recursions (involving gamma functions), but
the formulas are a little complicated, readers are referred to N.M. Temme,
<span class="emphasis"><em>On the numerical evaluation of the modified Bessel function of
the third kind</em></span>, Journal of Computational Physics, vol 19, 324
(1975). Note: Temme's series converge only for |μ| <= 1/2.
</p>
<p>
K<sub>ν</sub>(x) is then calculated from the forward recurrence, as is K<sub>ν+1</sub>(x). With
these two values and f<sub>ν</sub>, the Wronskian yields I<sub>ν</sub>(x) directly.
</p>
</div>
<table xmlns:rev="http://www.cs.rpi.edu/~gregod/boost/tools/doc/revision" width="100%"><tr>
<td align="left"></td>
<td align="right"><div class="copyright-footer">Copyright © 2006 , 2007, 2008, 2009 John Maddock, Paul A. Bristow,
Hubert Holin, Xiaogang Zhang, Bruno Lalande, Johan Råde, Gautam Sewani
and Thijs van den Berg<p>
Distributed under the Boost Software License, Version 1.0. (See accompanying
file LICENSE_1_0.txt or copy at <a href="http://www.boost.org/LICENSE_1_0.txt" target="_top">http://www.boost.org/LICENSE_1_0.txt</a>)
</p>
</div></td>
</tr></table>
<hr>
<div class="spirit-nav">
<a accesskey="p" href="bessel.html"><img src="../../../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../bessel.html"><img src="../../../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../../index.html"><img src="../../../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="sph_bessel.html"><img src="../../../../../../../../doc/src/images/next.png" alt="Next"></a>
</div>
</body>
</html>
