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.
 
 
 

160 lines
6.1 KiB

// <copyright file="MCMCSampler.cs" company="Math.NET">
// Math.NET Numerics, part of the Math.NET Project
// http://numerics.mathdotnet.com
// http://github.com/mathnet/mathnet-numerics
// http://mathnetnumerics.codeplex.com
//
// Copyright (c) 2009-2010 Math.NET
//
// Permission is hereby granted, free of charge, to any person
// obtaining a copy of this software and associated documentation
// files (the "Software"), to deal in the Software without
// restriction, including without limitation the rights to use,
// copy, modify, merge, publish, distribute, sublicense, and/or sell
// copies of the Software, and to permit persons to whom the
// Software is furnished to do so, subject to the following
// conditions:
//
// The above copyright notice and this permission notice shall be
// included in all copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
// OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
// NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
// HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
// WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
// FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
// OTHER DEALINGS IN THE SOFTWARE.
// </copyright>
namespace MathNet.Numerics.Statistics.Mcmc
{
using System;
/// <summary>
/// A method which samples datapoints from a proposal distribution. The implementation of this sampler
/// is stateless: no variables are saved between two calls to Sample. This proposal is different from
/// <seealso cref="LocalProposalSampler{T}"/> in that it doesn't take any parameters; it samples random
/// variables from the whole domain.
/// </summary>
/// <typeparam name="T">The type of the datapoints.</typeparam>
/// <returns>A sample from the proposal distribution.</returns>
public delegate T GlobalProposalSampler<out T>();
/// <summary>
/// A method which samples datapoints from a proposal distribution given an initial sample. The implementation
/// of this sampler is stateless: no variables are saved between two calls to Sample. This proposal is different from
/// <seealso cref="GlobalProposalSampler{T}"/> in that it samples locally around an initial point. In other words, it
/// makes a small local move rather than producing a global sample from the proposal.
/// </summary>
/// <typeparam name="T">The type of the datapoints.</typeparam>
/// <param name="init">The initial sample.</param>
/// <returns>A sample from the proposal distribution.</returns>
public delegate T LocalProposalSampler<T>(T init);
/// <summary>
/// A function which evaluates a density.
/// </summary>
/// <typeparam name="T">The type of data the distribution is over.</typeparam>
/// <param name="sample">The sample we want to evaluate the density for.</param>
public delegate double Density<in T>(T sample);
/// <summary>
/// A function which evaluates a log density.
/// </summary>
/// <typeparam name="T">The type of data the distribution is over.</typeparam>
/// <param name="sample">The sample we want to evaluate the log density for.</param>
public delegate double DensityLn<in T>(T sample);
/// <summary>
/// A function which evaluates the log of a transition kernel probability.
/// </summary>
/// <typeparam name="T">The type for the space over which this transition kernel is defined.</typeparam>
/// <param name="to">The new state in the transition.</param>
/// <param name="from">The previous state in the transition.</param>
/// <returns>The log probability of the transition.</returns>
public delegate double TransitionKernelLn<in T>(T to, T from);
/// <summary>
/// The interface which every sampler must implement.
/// </summary>
/// <typeparam name="T">The type of samples this sampler produces.</typeparam>
public abstract class McmcSampler<T>
{
/// <summary>
/// The random number generator for this class.
/// </summary>
private Random _randomNumberGenerator;
/// <summary>
/// Keeps track of the number of accepted samples.
/// </summary>
protected int Accepts;
/// <summary>
/// Keeps track of the number of calls to the proposal sampler.
/// </summary>
protected int Samples;
/// <summary>
/// Initializes a new instance of the <see cref="McmcSampler{T}"/> class.
/// </summary>
/// <remarks>Thread safe instances are two and half times slower than non-thread
/// safe classes.</remarks>
protected McmcSampler()
{
Accepts = 0;
Samples = 0;
RandomSource = new Random();
}
/// <summary>
/// Gets or sets the random number generator.
/// </summary>
/// <exception cref="ArgumentNullException">When the random number generator is null.</exception>
public Random RandomSource
{
get { return _randomNumberGenerator; }
set
{
if (value == null)
{
throw new ArgumentNullException();
}
_randomNumberGenerator = value;
}
}
/// <summary>
/// Returns one sample.
/// </summary>
public abstract T Sample();
/// <summary>
/// Returns a number of samples.
/// </summary>
/// <param name="n">The number of samples we want.</param>
/// <returns>An array of samples.</returns>
public virtual T[] Sample(int n)
{
T[] ret = new T[n];
for (int i = 0; i < n; i++)
{
ret[i] = Sample();
}
return ret;
}
/// <summary>
/// Gets the acceptance rate of the sampler.
/// </summary>
public double AcceptanceRate
{
get { return Accepts / (double)Samples; }
}
}
}