boost_1_45_0/libs/math/doc/sf_and_dist/distributions/laplace.qbk

[section:laplace_dist Laplace Distribution]

``#include <boost/math/distributions/laplace.hpp>``

   namespace boost{ namespace math{

   template <class RealType = double,
             class ``__Policy``   = ``__policy_class`` >
   class laplace_distribution;

   typedef laplace_distribution<> laplace;

   template <class RealType, class ``__Policy``>
   class laplace_distribution
   {
   public:
      typedef RealType value_type;
      typedef Policy   policy_type;
      // Construct:
      laplace_distribution(RealType location = 0, RealType scale = 1);
      // Accessors:
      RealType location()const;
      RealType scale()const;
   };

   }} // namespaces

Laplace distribution is the distribution of differences between two independent variates
with identical exponential distributions (Abramowitz and Stegun 1972, p. 930).
It is also called the double exponential distribution.

[/ Wikipedia definition is The difference between two independent identically distributed
exponential random variables is governed by a Laplace distribution.]

For location parameter [mu][space] and scale parameter [sigma][space] it is defined by the
probability density function:

[equation laplace_pdf]

The location and scale parameters are equivalent to the mean and
standard deviation of the normal or Gaussian distribution.

The following graph illustrates the effect of the 
parameters [mu][space] and [sigma][space] on the PDF. 
Note that the range of the random
variable remains \[-[infin],+[infin]\] irrespective of the value of the
location parameter:

[graph laplace_pdf]

[h4 Member Functions]

   laplace_distribution(RealType location = 0, RealType scale = 1);

Constructs a laplace distribution with location /location/ and
scale /scale/.

The location parameter is the same as the mean of the random variate.

The scale parameter is proportional to the standard deviation of the random variate.

Requires that the scale parameter is greater than zero, otherwise calls
__domain_error.

   RealType location()const;

Returns the /location/ parameter of this distribution.

   RealType scale()const;

Returns the /scale/ parameter of this distribution.

[h4 Non-member Accessors]

All the [link math_toolkit.dist.dist_ref.nmp usual non-member accessor functions] that are generic to all
distributions are supported: __usual_accessors.

The domain of the random variable is \[-[infin],+[infin]\].

[h4 Accuracy]

The laplace distribution is implemented in terms of the
standard library log and exp functions and as such should have very small errors.

[h4 Implementation]

In the following table [mu] is the location parameter of the distribution,
[sigma] is its scale parameter, /x/ is the random variate, /p/ is the probability
and its complement /q = 1-p/.

[table
[[Function][Implementation Notes]]
[[pdf][Using the relation: pdf = e[super -abs(x-[mu]) \/ [sigma]] \/ (2 * [sigma]) ]]
[[cdf][Using the relations:

x <  [mu] : p =  e[super (x-[mu])/[sigma] ] \/ [sigma]

x >= [mu] : p =  1 - e[super ([mu]-x)/[sigma] ] \/ [sigma]
]]
[[cdf complement][Using the relation:

-x <  [mu] : q =  e[super (-x-[mu])/[sigma] ] \/ [sigma]

-x >= [mu] : q =  1 - e[super ([mu]+x)/[sigma] ] \/ [sigma]
]]
[[quantile][Using the relations:

p <  0.5 : x = [mu] + [sigma] * log(2*p)

p >= 0.5 : x = [mu] - [sigma] * log(2-2*p)
]]
[[quantile from the complement][Using the relation:

q > 0.5: x = [mu] + [sigma]*log(2-2*q)

q <=0.5: x = [mu] - [sigma]*log( 2*q )
]]
[[mean][[mu]]]
[[variance][2 * [sigma][super 2] ]]
[[mode][[mu]]]
[[skewness][0]]
[[kurtosis][6]]
[[kurtosis excess][3]]
]

[h4 References]

* [@http://mathworld.wolfram.com/LaplaceDistribution.html Weisstein, Eric W. "Laplace Distribution."] From MathWorld--A Wolfram Web Resource.

* [@http://en.wikipedia.org/wiki/Laplace_distribution Laplace Distribution]

* M. Abramowitz and I. A. Stegun, Handbook of Mathematical Functions, 1972, p. 930.

[endsect][/section:laplace_dist laplace]

[/
  Copyright 2008, 2009 John Maddock, Paul A. Bristow and M.A. (Thijs) van den Berg.
  Distributed under the Boost Software License, Version 1.0.
  (See accompanying file LICENSE_1_0.txt or copy at
  http://www.boost.org/LICENSE_1_0.txt).
]