tsai
/
mathnet-numerics
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
Math.NET Numerics
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
17 Commits
25 Branches
121 Tags
187 MiB
 
 
 
Tree: 9bb9b575fd
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '9bb9b575fd'
${ noResults }
mathnet-numerics/Probability.html

482 lines
36 KiB
Raw Blame History

<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8"/>
<title>Probability Distributions
</title>
<meta name="viewport" content="width=device-width, initial-scale=1.0"/>
<meta name="description" content="Math.NET Numerics, providing methods and algorithms for numerical computations in science, engineering and every day use. .Net 4, .Net 3.5, SL5, Win8, WP8, PCL 47 and 136, Mono, Xamarin Android/iOS."/>
<meta name="author" content="Christoph Ruegg, Marcus Cuda, Jurgen Van Gael"/>
<script src="https://code.jquery.com/jquery-1.8.0.js"></script>
<script src="https://code.jquery.com/ui/1.8.23/jquery-ui.js"></script>
<script src="https://netdna.bootstrapcdn.com/twitter-bootstrap/2.2.1/js/bootstrap.min.js"></script>
<link href="https://netdna.bootstrapcdn.com/twitter-bootstrap/2.2.1/css/bootstrap-combined.min.css" rel="stylesheet"/>
<link type="text/css" rel="stylesheet" href="https://numerics.mathdotnet.com/content/style.css" />
<style>
#main table:not(.pre) {
border: 1px solid #dddddd;
max-width: 100%;
border-style: solid;
border-width: 1px;
border-color: gray;
border-collapse: collapse;
border-right-width: 1px;
border-bottom-width: 1px;
margin-top: 15px;
margin-bottom: 25px;
}
#main table:not(.pre) th, #main table:not(.pre) td {
border: 1px solid #dddddd;
padding: 6px;
}
#main table:not(.pre) th p, #main table:not(.pre) td p {
margin-bottom: 5px;
}
</style>
<script type="text/javascript" src="https://numerics.mathdotnet.com/content/tips.js"></script>
<!-- HTML5 shim, for IE6-8 support of HTML5 elements -->
<!--[if lt IE 9]>
<script src="https://oss.maxcdn.com/html5shiv/3.7.2/html5shiv.min.js"></script>
<![endif]-->
<script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.1/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
</head>
<body>
<div class="container">
<div class="masthead">
<ul class="nav nav-pills pull-right">
<li><a href="https://www.mathdotnet.com">Math.NET Project</a></li>
<li><a href="https://numerics.mathdotnet.com">Math.NET Numerics</a></li>
<li><a href="https://github.com/mathnet/mathnet-numerics">GitHub</a></li>
</ul>
<h3 class="muted">Math.NET Numerics</h3>
</div>
<hr />
<div class="row">
<div class="span9" id="main">
<h1><a name="Probability-Distributions" class="anchor" href="#Probability-Distributions">Probability Distributions</a></h1>
<p>Math.NET Numerics provides a wide range of probability distributions. Given the
distribution parameters they can be used to investigate their statistical properties
or to sample non-uniform random numbers.</p>
<p>All the distributions implement a common set of operations such as
evaluating the density (PDF) and the cumulative distribution (CDF)
at a given point, or to compute the mean, standard deviation and other properties.
Because it is often numerically more stable and faster to compute such statistical quantities
in the logarithmic domain, we also provide a selection of them in the log domain with the "Ln" suffix,
e.g. DensityLn for the logarithmic density.</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l"> 1: </span>
<span class="l"> 2: </span>
<span class="l"> 3: </span>
<span class="l"> 4: </span>
<span class="l"> 5: </span>
<span class="l"> 6: </span>
<span class="l"> 7: </span>
<span class="l"> 8: </span>
<span class="l"> 9: </span>
<span class="l">10: </span>
<span class="l">11: </span>
<span class="l">12: </span>
<span class="l">13: </span>
<span class="l">14: </span>
<span class="l">15: </span>
<span class="l">16: </span>
<span class="l">17: </span>
<span class="l">18: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="csharp"><span class="k">using</span> MathNet.Numerics.Distributions;
<span class="k">using</span> MathNet.Numerics.Random;
<span class="c">// create a parametrized distribution instance</span>
<span class="k">var</span> gamma <span class="o">=</span> <span class="k">new</span> Gamma(<span class="n">2.0</span>, <span class="n">1.5</span>);
<span class="c">// distribution properties</span>
<span class="k">double</span> mean <span class="o">=</span> gamma.Mean;
<span class="k">double</span> variance <span class="o">=</span> gamma.Variance;
<span class="k">double</span> entropy <span class="o">=</span> gamma.Entropy;
<span class="c">// distribution functions</span>
<span class="k">double</span> a <span class="o">=</span> gamma.Density(<span class="n">2.3</span>); <span class="c">// PDF</span>
<span class="k">double</span> b <span class="o">=</span> gamma.DensityLn(<span class="n">2.3</span>); <span class="c">// ln(PDF)</span>
<span class="k">double</span> c <span class="o">=</span> gamma.CumulativeDistribution(<span class="n">0.7</span>); <span class="c">// CDF</span>
<span class="c">// non-uniform number sampling</span>
<span class="k">double</span> randomSample <span class="o">=</span> gamma.Sample();
</code></pre></td></tr></table>
<p>Both probability functions and sampling are also available as static functions
for simpler usage scenarios:</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
<span class="l">3: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="csharp"><span class="c">// distribution parameters must be passed as arguments</span>
<span class="k">double</span> a<span class="n">2</span> <span class="o">=</span> Gamma.PDF(<span class="n">2.0</span>, <span class="n">1.5</span>, <span class="n">2.3</span>);
<span class="k">double</span> randomSample<span class="n">2</span> <span class="o">=</span> Gamma.Sample(<span class="n">2.0</span>, <span class="n">1.5</span>);
</code></pre></td></tr></table>
<div style="overflow:auto;">
<div style="float: left; width: 50%;">
<h3><a name="Continuous-Distributions" class="anchor" href="#Continuous-Distributions">Continuous Distributions</a></h3>
<ul>
<li><a href="https://en.wikipedia.org/wiki/Uniform_distribution_%28continuous%29">Continuous Uniform</a></li>
<li><a href="https://en.wikipedia.org/wiki/Normal_distribution">Normal</a></li>
<li><a href="https://en.wikipedia.org/wiki/Log-normal_distribution">Log Normal</a></li>
<li><a href="https://en.wikipedia.org/wiki/Beta_distribution">Beta</a></li>
<li><a href="https://en.wikipedia.org/wiki/cauchy_distribution">Cauchy</a> (Cauchy-Lorentz)</li>
<li><a href="https://en.wikipedia.org/wiki/Chi_distribution">Chi</a></li>
<li><a href="https://en.wikipedia.org/wiki/Chi-square_distribution">Chi Squared</a></li>
<li><a href="https://en.wikipedia.org/wiki/Erlang_distribution">Erlang</a></li>
<li><a href="https://en.wikipedia.org/wiki/exponential_distribution">Exponential</a></li>
<li><a href="https://en.wikipedia.org/wiki/F-distribution">Fisher-Snedecor</a> (F-Distribution)</li>
<li><a href="https://en.wikipedia.org/wiki/Gamma_distribution">Gamma</a></li>
<li><a href="https://en.wikipedia.org/wiki/inverse-gamma_distribution">Inverse Gamma</a></li>
<li><a href="https://en.wikipedia.org/wiki/Laplace_distribution">Laplace</a></li>
<li><a href="https://en.wikipedia.org/wiki/Pareto_distribution">Pareto</a></li>
<li><a href="https://en.wikipedia.org/wiki/Rayleigh_distribution">Rayleigh</a></li>
<li><a href="https://en.wikipedia.org/wiki/Stable_distribution">Stable</a></li>
<li><a href="https://en.wikipedia.org/wiki/Student%27s_t-distribution">Stundent-T</a></li>
<li><a href="https://en.wikipedia.org/wiki/Weibull_distribution">Weibull</a></li>
<li><a href="https://en.wikipedia.org/wiki/Triangular_distribution">Triangular</a></li>
</ul>
</div>
<div style="float: right; width: 50%;">
<h3><a name="Discrete-Distributions" class="anchor" href="#Discrete-Distributions">Discrete Distributions</a></h3>
<ul>
<li><a href="https://en.wikipedia.org/wiki/Uniform_distribution_%28discrete%29">Discrete Uniform</a></li>
<li><a href="https://en.wikipedia.org/wiki/Bernoulli_distribution">Bernoulli</a></li>
<li><a href="https://en.wikipedia.org/wiki/Binomial_distribution">Binomial</a></li>
<li><a href="https://en.wikipedia.org/wiki/Negative_binomial_distribution">Negative Binomial</a></li>
<li><a href="https://en.wikipedia.org/wiki/geometric_distribution">Geometric</a></li>
<li><a href="https://en.wikipedia.org/wiki/Hypergeometric_distribution">Hypergeometric</a></li>
<li><a href="https://en.wikipedia.org/wiki/Poisson_distribution">Poisson</a></li>
<li><a href="https://en.wikipedia.org/wiki/Categorical_distribution">Categorical</a></li>
<li><a href="https://en.wikipedia.org/wiki/Conway%E2%80%93Maxwell%E2%80%93Poisson_distribution">Conway-Maxwell-Poisson</a></li>
<li><a href="https://en.wikipedia.org/wiki/Zipf%27s_law">Zipf</a></li>
</ul>
<h3><a name="Multivariate-Distributions" class="anchor" href="#Multivariate-Distributions">Multivariate Distributions</a></h3>
<ul>
<li><a href="https://en.wikipedia.org/wiki/Dirichlet_distribution">Dirichlet</a></li>
<li><a href="https://en.wikipedia.org/wiki/Inverse-Wishart_distribution">Inverse Wishart</a></li>
<li><a href="https://en.wikipedia.org/wiki/Matrix_normal_distribution">Matrix Normal</a></li>
<li><a href="https://en.wikipedia.org/wiki/Multinomial_distribution">Multinomial</a></li>
<li><a href="https://en.wikipedia.org/wiki/Normal-gamma_distribution">Normal Gamma</a></li>
<li><a href="https://en.wikipedia.org/wiki/Wishart_distribution">Wishart</a></li>
</ul>
</div>
</div>
<h2><a name="Distribution-Parameters" class="anchor" href="#Distribution-Parameters">Distribution Parameters</a></h2>
<p>There are many ways to parametrize a distribution in the literature. When using the
default constructor, read carefully which parameters it requires. For distributions where
multiple ways are common there are also static methods, so you can use the one that fits best.
For example, a normal distribution is usually parametrized with mean and standard deviation,
but if you'd rather use mean and precision:</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="csharp"><span class="k">var</span> normal <span class="o">=</span> Normal.WithMeanPrecision(<span class="n">0.0</span>, <span class="n">0.5</span>);
</code></pre></td></tr></table>
<p>Since probability distributions can also be sampled to generate random numbers
with the configured distribution, all constructors optionally accept a random generator
as last argument.</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
<span class="l">3: </span>
<span class="l">4: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="csharp"><span class="k">var</span> gamma<span class="n">2</span> <span class="o">=</span> <span class="k">new</span> Gamma(<span class="n">2.0</span>, <span class="n">1.5</span>, <span class="k">new</span> MersenneTwister());
<span class="c">// the random generator can also be replaced on an existing instance</span>
gamma<span class="n">2</span>.RandomSource <span class="o">=</span> <span class="k">new</span> Mrg<span class="n">32</span>k<span class="n">3</span>a();
</code></pre></td></tr></table>
<p>A few more examples, this time in F#:</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
<span class="l">3: </span>
<span class="l">4: </span>
<span class="l">5: </span>
<span class="l">6: </span>
<span class="l">7: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="fsharp"><span class="c">// some probability distributions</span>
<span class="k">let</span> <span class="i">normal</span> <span class="o">=</span> <span class="i">Normal</span><span class="o">.</span><span class="i">WithMeanVariance</span>(<span class="n">3.0</span>, <span class="n">1.5</span>, <span class="i">a</span>)
<span class="k">let</span> <span class="i">exponential</span> <span class="o">=</span> <span class="i">Exponential</span>(<span class="n">2.4</span>)
<span class="k">let</span> <span class="i">gamma</span> <span class="o">=</span> <span class="i">Gamma</span>(<span class="n">2.0</span>, <span class="n">1.5</span>, <span class="i">Random</span><span class="o">.</span><span class="i">crypto</span>())
<span class="k">let</span> <span class="i">cauchy</span> <span class="o">=</span> <span class="i">Cauchy</span>(<span class="n">0.0</span>, <span class="n">1.0</span>, <span class="i">Random</span><span class="o">.</span><span class="i">mrg32k3aWith</span> <span class="n">10</span> <span class="k">false</span>)
<span class="k">let</span> <span class="i">poisson</span> <span class="o">=</span> <span class="i">Poisson</span>(<span class="n">3.0</span>)
<span class="k">let</span> <span class="i">geometric</span> <span class="o">=</span> <span class="i">Geometric</span>(<span class="n">0.8</span>, <span class="i">Random</span><span class="o">.</span><span class="i">system</span>())
</code></pre></td>
</tr>
</table>
<p>Some of the distributions also have routines for maximum-likelihood parameter
estimation from a set of samples:</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
<span class="l">3: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="fsharp"><span class="k">let</span> <span class="i">estimation</span> <span class="o">=</span> <span class="i">LogNormal</span><span class="o">.</span><span class="i">Estimate</span>([| <span class="n">2.0</span>; <span class="n">1.5</span>; <span class="n">2.1</span>; <span class="n">1.2</span>; <span class="n">3.0</span>; <span class="n">2.4</span>; <span class="n">1.8</span> |])
<span class="k">let</span> <span class="i">mean</span>, <span class="i">variance</span> <span class="o">=</span> <span class="i">estimation</span><span class="o">.</span><span class="i">Mean</span>, <span class="i">estimation</span><span class="o">.</span><span class="i">Variance</span>
<span class="k">let</span> <span class="i">moreSamples</span> <span class="o">=</span> <span class="i">estimation</span><span class="o">.</span><span class="i">Samples</span>() <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">take</span> <span class="n">10</span> <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">toArray</span>
</code></pre></td>
</tr>
</table>
<p>or in C#:</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
<span class="l">3: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="csharp">LogNormal estimation <span class="o">=</span> LogNormal.Estimate(<span class="k">new</span> [] {<span class="n">2.0</span>, <span class="n">1.5</span>, <span class="n">2.1</span>, <span class="n">1.2</span>, <span class="n">3.0</span>, <span class="n">2.4</span>, <span class="n">1.8</span>});
<span class="k">double</span> mean <span class="o">=</span> estimation.Mean, variance <span class="o">=</span> estimation.Variance;
<span class="k">double</span>[] moreSamples <span class="o">=</span> estimation.Samples().Take(<span class="n">10</span>).ToArray();
</code></pre></td></tr></table>
<h2><a name="Sampling-a-Probability-Distribution" class="anchor" href="#Sampling-a-Probability-Distribution">Sampling a Probability Distribution</a></h2>
<p>Each distribution provides methods to generate random numbers from that distribution.
These random variate generators work by accessing the distribution's member RandomSource
to provide uniform random numbers. By default, this member is an instance of System.Random
but one can easily replace this with more sophisticated random number generators from
<code>MathNet.Numerics.Random</code> (see <a href="Random.html">Random Numbers</a> for details).</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l"> 1: </span>
<span class="l"> 2: </span>
<span class="l"> 3: </span>
<span class="l"> 4: </span>
<span class="l"> 5: </span>
<span class="l"> 6: </span>
<span class="l"> 7: </span>
<span class="l"> 8: </span>
<span class="l"> 9: </span>
<span class="l">10: </span>
<span class="l">11: </span>
<span class="l">12: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="fsharp"><span class="c">// sample some random numbers from these distributions</span>
<span class="c">// continuous distributions sample to floating-point numbers:</span>
<span class="k">let</span> <span class="i">continuous</span> <span class="o">=</span>
[ <span class="k">yield</span> <span class="i">normal</span><span class="o">.</span><span class="i">Sample</span>()
<span class="k">yield</span> <span class="i">exponential</span><span class="o">.</span><span class="i">Sample</span>()
<span class="k">yield!</span> <span class="i">gamma</span><span class="o">.</span><span class="i">Samples</span>() <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">take</span> <span class="n">10</span> ]
<span class="c">// discrete distributions on the other hand sample to integers:</span>
<span class="k">let</span> <span class="i">discrete</span> <span class="o">=</span>
[ <span class="i">poisson</span><span class="o">.</span><span class="i">Sample</span>()
<span class="i">poisson</span><span class="o">.</span><span class="i">Sample</span>()
<span class="i">geometric</span><span class="o">.</span><span class="i">Sample</span>() ]
</code></pre></td>
</tr>
</table>
<p>Instead of creating a distribution object we can also sample directly with static functions.
Note that no intermediate value caching is possible this way and parameters must be validated on each call.</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
<span class="l">3: </span>
<span class="l">4: </span>
<span class="l">5: </span>
<span class="l">6: </span>
<span class="l">7: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="fsharp"><span class="c">// using the default number generator (SystemRandomSource.Default)</span>
<span class="k">let</span> <span class="i">w</span> <span class="o">=</span> <span class="i">Rayleigh</span><span class="o">.</span><span class="i">Sample</span>(<span class="n">1.5</span>)
<span class="k">let</span> <span class="i">x</span> <span class="o">=</span> <span class="i">Hypergeometric</span><span class="o">.</span><span class="i">Sample</span>(<span class="n">100</span>, <span class="n">20</span>, <span class="n">5</span>)
<span class="c">// or by manually providing the uniform random number generator</span>
<span class="k">let</span> <span class="i">u</span> <span class="o">=</span> <span class="i">Normal</span><span class="o">.</span><span class="i">Sample</span>(<span class="i">Random</span><span class="o">.</span><span class="i">system</span>(), <span class="n">2.0</span>, <span class="n">4.0</span>)
<span class="k">let</span> <span class="i">v</span> <span class="o">=</span> <span class="i">Laplace</span><span class="o">.</span><span class="i">Samples</span>(<span class="i">Random</span><span class="o">.</span><span class="i">mersenneTwister</span>(), <span class="n">1.0</span>, <span class="n">3.0</span>) <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">take</span> <span class="n">100</span> <span class="o">|&gt;</span> <span class="i">List</span><span class="o">.</span><span class="i">ofSeq</span>
</code></pre></td>
</tr>
</table>
<p>If you need to sample not just one or two values but a large number of them,
there are routines that either fill an existing array or return an enumerable.
The variant that fills an array is generally the fastest. Routines to sample
more than one value use the plural form <code>Samples</code> instead of <code>Sample</code>.</p>
<p>Let's sample 100'000 values from a laplace distribution with mean 1.0 and scale 2.0 in C#:</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="csharp"><span class="k">var</span> samples <span class="o">=</span> <span class="k">new</span> <span class="k">double</span>[<span class="n">100000</span>];
Laplace.Samples(samples, <span class="n">1.0</span>, <span class="n">2.0</span>);
</code></pre></td></tr></table>
<p>Let's do some random walks in F# (TODO: Graph):</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l">1: </span>
<span class="l">2: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="fsharp"><span class="i">Seq</span><span class="o">.</span><span class="i">scan</span> (<span class="o">+</span>) <span class="n">0.0</span> (<span class="i">Normal</span><span class="o">.</span><span class="i">Samples</span>(<span class="n">0.0</span>, <span class="n">1.0</span>)) <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">take</span> <span class="n">10</span> <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">toArray</span>
<span class="i">Seq</span><span class="o">.</span><span class="i">scan</span> (<span class="o">+</span>) <span class="n">0.0</span> (<span class="i">Cauchy</span><span class="o">.</span><span class="i">Samples</span>(<span class="n">0.0</span>, <span class="n">1.0</span>)) <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">take</span> <span class="n">10</span> <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">toArray</span>
</code></pre></td>
</tr>
</table>
<h2><a name="Distribution-Functions-and-Properties" class="anchor" href="#Distribution-Functions-and-Properties">Distribution Functions and Properties</a></h2>
<p>Distributions can not just be used to generate non-uniform random samples.
Once parametrized they can compute a variety of distribution properties
or evaluate distribution functions. Because it is often numerically more stable
and faster to compute and work with such quantities in the logarithmic domain,
some of them are also available with the <code>Ln</code>-suffix.</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l"> 1: </span>
<span class="l"> 2: </span>
<span class="l"> 3: </span>
<span class="l"> 4: </span>
<span class="l"> 5: </span>
<span class="l"> 6: </span>
<span class="l"> 7: </span>
<span class="l"> 8: </span>
<span class="l"> 9: </span>
<span class="l">10: </span>
<span class="l">11: </span>
<span class="l">12: </span>
<span class="l">13: </span>
<span class="l">14: </span>
<span class="l">15: </span>
<span class="l">16: </span>
<span class="l">17: </span>
<span class="l">18: </span>
<span class="l">19: </span>
<span class="l">20: </span>
<span class="l">21: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="fsharp"><span class="c">// distribution properties of the gamma we&#39;ve configured above</span>
<span class="k">let</span> <span class="i">gammaStats</span> <span class="o">=</span>
( <span class="i">gamma</span><span class="o">.</span><span class="i">Mean</span>,
<span class="i">gamma</span><span class="o">.</span><span class="i">Variance</span>,
<span class="i">gamma</span><span class="o">.</span><span class="i">StdDev</span>,
<span class="i">gamma</span><span class="o">.</span><span class="i">Entropy</span>,
<span class="i">gamma</span><span class="o">.</span><span class="i">Skewness</span>,
<span class="i">gamma</span><span class="o">.</span><span class="i">Mode</span> )
<span class="c">// probability distribution functions of the normal we&#39;ve configured above.</span>
<span class="k">let</span> <span class="i">nd</span> <span class="o">=</span> <span class="i">normal</span><span class="o">.</span><span class="i">Density</span>(<span class="n">4.0</span>) <span class="c">(* PDF *)</span>
<span class="k">let</span> <span class="i">ndLn</span> <span class="o">=</span> <span class="i">normal</span><span class="o">.</span><span class="i">DensityLn</span>(<span class="n">4.0</span>) <span class="c">(* ln(PDF) *)</span>
<span class="k">let</span> <span class="i">nc</span> <span class="o">=</span> <span class="i">normal</span><span class="o">.</span><span class="i">CumulativeDistribution</span>(<span class="n">4.0</span>) <span class="c">(* CDF *)</span>
<span class="k">let</span> <span class="i">nic</span> <span class="o">=</span> <span class="i">normal</span><span class="o">.</span><span class="i">InverseCumulativeDistribution</span>(<span class="n">0.7</span>) <span class="c">(* CDF^(-1) *)</span>
<span class="c">// Distribution functions can also be evaluated without creating an object,</span>
<span class="c">// but then you have to pass in the distribution parameters as first arguments:</span>
<span class="k">let</span> <span class="i">nd2</span> <span class="o">=</span> <span class="i">Normal</span><span class="o">.</span><span class="i">PDF</span>(<span class="n">3.0</span>, <span class="i">sqrt</span> <span class="n">1.5</span>, <span class="n">4.0</span>)
<span class="k">let</span> <span class="i">ndLn2</span> <span class="o">=</span> <span class="i">Normal</span><span class="o">.</span><span class="i">PDFLn</span>(<span class="n">3.0</span>, <span class="i">sqrt</span> <span class="n">1.5</span>, <span class="n">4.0</span>)
<span class="k">let</span> <span class="i">nc2</span> <span class="o">=</span> <span class="i">Normal</span><span class="o">.</span><span class="i">CDF</span>(<span class="n">3.0</span>, <span class="i">sqrt</span> <span class="n">1.5</span>, <span class="n">4.0</span>)
<span class="k">let</span> <span class="i">nic2</span> <span class="o">=</span> <span class="i">Normal</span><span class="o">.</span><span class="i">InvCDF</span>(<span class="n">3.0</span>, <span class="i">sqrt</span> <span class="n">1.5</span>, <span class="n">0.7</span>)
</code></pre></td>
</tr>
</table>
<h2><a name="Composing-Distributions" class="anchor" href="#Composing-Distributions">Composing Distributions</a></h2>
<p>Specifically for F# there is also a <code>Sample</code> module that allows a somewhat more functional
view on distribution sampling functions by having the random source passed in as last argument.
This way they can be composed and transformed arbitrarily if curried:</p>
<table class="pre"><tr><td class="lines"><pre class="fssnip"><span class="l"> 1: </span>
<span class="l"> 2: </span>
<span class="l"> 3: </span>
<span class="l"> 4: </span>
<span class="l"> 5: </span>
<span class="l"> 6: </span>
<span class="l"> 7: </span>
<span class="l"> 8: </span>
<span class="l"> 9: </span>
<span class="l">10: </span>
<span class="l">11: </span>
<span class="l">12: </span>
<span class="l">13: </span>
<span class="l">14: </span>
<span class="l">15: </span>
<span class="l">16: </span>
</pre></td>
<td class="snippet"><pre class="fssnip highlighted"><code lang="fsharp"><span class="c">/// Transform a sample from a distribution</span>
<span class="k">let</span> <span class="i">s1</span> <span class="i">rng</span> <span class="o">=</span> <span class="i">tanh</span> (<span class="i">Sample</span><span class="o">.</span><span class="i">normal</span> <span class="n">2.0</span> <span class="n">0.5</span> <span class="i">rng</span>)
<span class="c">/// But we really want to transform the function, not the resulting sample:</span>
<span class="k">let</span> <span class="i">s1f</span> <span class="i">rng</span> <span class="o">=</span> <span class="i">Sample</span><span class="o">.</span><span class="i">map</span> <span class="i">tanh</span> (<span class="i">Sample</span><span class="o">.</span><span class="i">normal</span> <span class="n">2.0</span> <span class="n">0.5</span>) <span class="i">rng</span>
<span class="c">/// Exactly the same also works with functions generating full sequences</span>
<span class="k">let</span> <span class="i">s1s</span> <span class="i">rng</span> <span class="o">=</span> <span class="i">Sample</span><span class="o">.</span><span class="i">mapSeq</span> <span class="i">tanh</span> (<span class="i">Sample</span><span class="o">.</span><span class="i">normalSeq</span> <span class="n">2.0</span> <span class="n">0.5</span>) <span class="i">rng</span>
<span class="c">/// Now with multiple distributions, e.g. their product:</span>
<span class="k">let</span> <span class="i">s2</span> <span class="i">rng</span> <span class="o">=</span> (<span class="i">Sample</span><span class="o">.</span><span class="i">normal</span> <span class="n">2.0</span> <span class="n">1.5</span> <span class="i">rng</span>) <span class="o">*</span> (<span class="i">Sample</span><span class="o">.</span><span class="i">cauchy</span> <span class="n">2.0</span> <span class="n">0.5</span> <span class="i">rng</span>)
<span class="k">let</span> <span class="i">s2f</span> <span class="i">rng</span> <span class="o">=</span> <span class="i">Sample</span><span class="o">.</span><span class="i">map2</span> <span class="o">(*)</span> (<span class="i">Sample</span><span class="o">.</span><span class="i">normal</span> <span class="n">2.0</span> <span class="n">1.5</span>) (<span class="i">Sample</span><span class="o">.</span><span class="i">cauchy</span> <span class="n">2.0</span> <span class="n">0.5</span>) <span class="i">rng</span>
<span class="k">let</span> <span class="i">s2s</span> <span class="i">rng</span> <span class="o">=</span> <span class="i">Sample</span><span class="o">.</span><span class="i">mapSeq2</span> <span class="o">(*)</span> (<span class="i">Sample</span><span class="o">.</span><span class="i">normalSeq</span> <span class="n">2.0</span> <span class="n">1.5</span>) (<span class="i">Sample</span><span class="o">.</span><span class="i">cauchySeq</span> <span class="n">2.0</span> <span class="n">0.5</span>) <span class="i">rng</span>
<span class="c">// Taking some samples from the composed function</span>
<span class="i">Seq</span><span class="o">.</span><span class="i">take</span> <span class="n">10</span> (<span class="i">s2s</span> (<span class="i">Random</span><span class="o">.</span><span class="i">system</span>())) <span class="o">|&gt;</span> <span class="i">Seq</span><span class="o">.</span><span class="i">toArray</span>
</code></pre></td>
</tr>
</table>
</div>
<div class="span3">
<ul class="nav nav-list" id="menu">
<li class="nav-header">Math.NET Numerics</li>
<li><a href="https://numerics.mathdotnet.com/Packages.html">NuGet & Binaries</a></li>
<li><a href="https://numerics.mathdotnet.com/ReleaseNotes.html">Release Notes</a></li>
<li><a href="https://numerics.mathdotnet.com/License.html">MIT/X11 License</a></li>
<li><a href="https://numerics.mathdotnet.com/Compatibility.html">Platform Support</a></li>
<li><a href="https://numerics.mathdotnet.com/api/">Class Reference</a></li>
<li><a href="https://github.com/mathnet/mathnet-numerics/issues">Issues & Bugs</a></li>
<li><a href="https://numerics.mathdotnet.com/Users.html">Who is using Math.NET?</a></li>
<li class="nav-header">Contributing</li>
<li><a href="https://numerics.mathdotnet.com/Contributors.html">Contributors</a></li>
<li><a href="https://numerics.mathdotnet.com/Contributing.html">Contributing</a></li>
<li><a href="https://numerics.mathdotnet.com/Build.html">Build & Tools</a></li>
<li><a href="http://feedback.mathdotnet.com/forums/2060-math-net-numerics">Your Ideas</a></li>
<li class="nav-header">Getting Help</li>
<li><a href="https://discuss.mathdotnet.com/c/numerics">Discuss</a></li>
<li><a href="https://stackoverflow.com/questions/tagged/mathdotnet">Stack Overflow</a></li>
<li class="nav-header">Getting Started</li>
<li><a href="https://numerics.mathdotnet.com/">Getting started</a></li>
<li><a href="https://numerics.mathdotnet.com/Constants.html">Constants</a></li>
<li>Floating-Point Numbers</li>
<li>Arbitrary Precision Numbers</li>
<li>Complex Numbers</li>
<li><a href="https://numerics.mathdotnet.com/Matrix.html">Matrices and Vectors</a></li>
<li><a href="https://numerics.mathdotnet.com/Euclid.html">Euclid & Number Theory</a></li>
<li>Combinatorics</li>
<li class="nav-header">Evaluation</li>
<li><a href="https://numerics.mathdotnet.com/Functions.html">Special Functions</a></li>
<li>Differentiation</li>
<li><a href="https://numerics.mathdotnet.com/Integration.html">Integration</a></li>
<li class="nav-header">Statistics/Probability</li>
<li><a href="https://numerics.mathdotnet.com/DescriptiveStatistics.html">Descriptive Statistics</a></li>
<li><a href="https://numerics.mathdotnet.com/Probability.html">Probability Distributions</a></li>
<li class="nav-header">Generation</li>
<li><a href="https://numerics.mathdotnet.com/Generate.html">Generating Data</a></li>
<li><a href="https://numerics.mathdotnet.com/Random.html">Random Numbers</a></li>
<li class="nav-header">Transformation</li>
<li>Fourier Transform (FFT)</li>
<li>Filtering & DSP</li>
<li>Window Functions</li>
<li class="nav-header">Solving Equations</li>
<li><a href="https://numerics.mathdotnet.com/LinearEquations.html">Linear Equation Systems</a></li>
<li>Nonlinear Root Finding</li>
<li class="nav-header">Optimization</li>
<li>Linear Least Squares</li>
<li>Nonlinear Optimization</li>
<li><a href="https://numerics.mathdotnet.com/Distance.html">Distance Metrics</a></li>
<li class="nav-header">Curve Fitting</li>
<li><a href="https://numerics.mathdotnet.com/Regression.html">Regression</a></li>
<li>Interpolation</li>
<li>Fourier Approximation</li>
<li class="nav-header">Native Providers</li>
<li><a href="https://numerics.mathdotnet.com/MKL.html">Intel MKL</a></li>
<li class="nav-header">Working Together</li>
<li><a href="https://numerics.mathdotnet.com/CSV.html">Delimited Text Files (CSV)</a></li>
<li><a href="https://numerics.mathdotnet.com/MatrixMarket.html">NIST MatrixMarket</a></li>
<li><a href="https://numerics.mathdotnet.com/MatlabFiles.html">MATLAB</a></li>
<li><a href="https://numerics.mathdotnet.com/IFSharpNotebook.html">IF# Notebook</a></li>
<li>FsLab & Deedle</li>
<li>Microsoft Excel</li>
<li>numl.net machine learning</li>
<li>R-project</li>
</ul>
</div>
</div>
</div>
</body>
</html>