/// <summary>
/// Creates a new Generalized Linear Regression Model.
/// </summary>
///
/// <param name="function">The link function to use.</param>
/// <param name="coefficients">The coefficient vector.</param>
/// <param name="standardErrors">The standard error vector.</param>
///
public GeneralizedLinearRegression(ILinkFunction function,
double[] coefficients, double[] standardErrors)
{
this.linkFunction = function;
this.coefficients = coefficients;
this.standardErrors = standardErrors;
}
/// <summary>
/// Creates a new Generalized Linear Regression Model.
/// </summary>
///
/// <param name="function">The link function to use.</param>
///
public GeneralizedLinearRegression(ILinkFunction function)
{
this.linear = new MultipleLinearRegression();
this.linkFunction = function;
this.NumberOfOutputs = 1;
this.NumberOfInputs = 1;
}
public static ILinkFunction Get(LinkFunctionType link)
{
ILinkFunction result = null;
switch (link)
{
case LinkFunctionType.Absolute: result = new AbsoluteLinkFunction(); break;
case LinkFunctionType.Cauchit: result = new CauchitLinkFunction(); break;
case LinkFunctionType.Identity: result = new IdentityLinkFunction(); break;
case LinkFunctionType.Inverse: result = new InverseLinkFunction(); break;
case LinkFunctionType.InverseSquared: result = new InverseSquaredLinkFunction(); break;
case LinkFunctionType.Logit: result = new LogitLinkFunction(); break;
case LinkFunctionType.Log: result = new LogLinkFunction(); break;
case LinkFunctionType.LogLog: result = new LogLogLinkFunction(); break;
case LinkFunctionType.Probit: result = new ProbitLinkFunction(); break;
case LinkFunctionType.Sin: result = new SinLinkFunction(); break;
case LinkFunctionType.Threshold: result = new ThresholdLinkFunction(); break;
}
return(result);
}
/// <summary>
/// Constructs a <see cref="GaussianDistribution"/>.
/// </summary>
/// <param name="mean">
/// The mean of the distribution. Defaults to zero.
/// </param>
/// <param name="standardDeviation">
/// The standard deviation of the distribution. Defaults to 1.0.
/// </param>
/// <param name="linkFunction">
/// The link function.
/// </param>
/// <param name="random">
/// A random number generator for the distribution. Defaults to <see cref="Random"/>.
/// </param>
public GaussianDistribution([NotNull] ILinkFunction linkFunction, double mean = 0.0, double standardDeviation = 1.0, [CanBeNull] Random random = null)
{
Mean = mean;
StandardDeviation = standardDeviation;
LinkFunction = linkFunction;
_random = random ?? new Random();
}
public GeneralizedLinearRegression(ILinkFunction function,
double[] coefficients, double[] standardErrors)
: this()
{
this.linkFunction = function;
this.Weights = coefficients.Get(1, 0);
this.StandardErrors = standardErrors;
}
/// <summary>
/// Creates a new Generalized Linear Regression Model.
/// </summary>
///
/// <param name="function">The link function to use.</param>
/// <param name="inputs">The number of input variables for the model.</param>
///
public GeneralizedLinearRegression(ILinkFunction function, int inputs)
{
this.linkFunction = function;
this.coefficients = new double[inputs + 1];
this.standardErrors = new double[inputs + 1];
this.NumberOfInputs = inputs;
this.NumberOfOutputs = 1;
}
public GlmIrls(GlmDistributionFamily distribution, ILinkFunction linkFunc, double[,] A, double[] b)
: base(distribution, linkFunc, null, null, null)
{
this.A = new SparseMatrix(A);
this.b = new SparseVector(b);
this.At = this.A.Transpose();
this.mStats = new Statistics.GlmStatistics(A.GetLength(1), b.Length);
}
public Glm(GlmDistributionFamily distribution, double[][] A, double[] b, SingleTrajectoryContinuousSolver solver)
{
this.solver = solver;
this.mDistributionFamily = distribution;
this.mLinkFunc = GetLinkFunction(distribution);
this.A = A;
this.b = b;
this.mStats = new Statistics.GlmStatistics(A[0].Length, b.Length);
}
/// <summary>
/// Creates a new GeneralizedLinearRegression that is a copy of the current instance.
/// </summary>
///
public object Clone()
{
ILinkFunction function = (ILinkFunction)linkFunction.Clone();
var regression = new GeneralizedLinearRegression(function, coefficients.Length);
regression.coefficients = (double[])this.coefficients.Clone();
regression.standardErrors = (double[])this.standardErrors.Clone();
return(regression);
}
/// <summary>
/// Constructs a <see cref="PoissonDistribution"/>.
/// </summary>
/// <param name="linkFunction">
/// The link function.
/// </param>
/// <param name="mean">
/// A <paramref name="mean"/> value for the distribution. Defaults to 1.0.
/// </param>
/// <param name="random">
/// A random number generator for the distribution. Defaults to <see cref="Random"/>.
/// </param>
public PoissonDistribution([NotNull] ILinkFunction linkFunction, double mean = 1.0, [CanBeNull] Random random = default)
{
if (mean <= 0)
{
throw new ArgumentOutOfRangeException(nameof(mean), "Argument must be greater than zero.");
}
Mean = mean;
_random = random ?? new Random();
LinkFunction = linkFunction;
_smallSampleEnumerator = SmallSamples();
_largeSampleEnumerator = LargeSamples();
}
public Glm(GlmDistributionFamily distribution, double[,] A, double[] b, SingleTrajectoryContinuousSolver solver, int maxIters = -1)
{
this.solver = solver;
this.mDistributionFamily = distribution;
this.mLinkFunc = GetLinkFunction(distribution);
this.A = new double[A.GetLength(0)][];
for (int i = 0; i < A.GetLength(0); i++)
{
this.A[i] = new double[A.GetLength(1)];
for (int j = 0; j < A.GetLength(1); j++)
{
this.A[i][j] = A[i, j];
}
}
this.b = b;
if (maxIters > 0)
{
this.mMaxIters = maxIters;
}
this.mStats = new Statistics.GlmStatistics(A.GetLength(1), b.Length);
}
public Glm(GlmDistributionFamily distribution)
{
this.mLinkFunc = GetLinkFunction(distribution);
this.mDistributionFamily = distribution;
}
//---------------------------------------------
#region Constructor
/// <summary>
/// Creates a new Generalized Linear Regression Model.
/// </summary>
///
/// <param name="function">The link function to use.</param>
/// <param name="inputs">The number of input variables for the model.</param>
///
public GeneralizedLinearRegression(ILinkFunction function, int inputs)
{
this.linkFunction = function;
this.coefficients = new double[inputs + 1];
this.standardErrors = new double[inputs + 1];
}
/// <summary>
/// Creates a new Generalized Linear Regression Model.
/// </summary>
///
/// <param name="function">The link function to use.</param>
/// <param name="inputs">The number of input variables for the model.</param>
/// <param name="intercept">The starting intercept value. Default is 0.</param>
///
public GeneralizedLinearRegression(ILinkFunction function, int inputs, double intercept)
: this(function, inputs)
{
this.coefficients[0] = intercept;
}
public GeneralizedLinearRegression(ILinkFunction function, int inputs)
: this(function)
{
this.NumberOfInputs = inputs;
}
/// <summary>
/// Creates a new Generalized Linear Regression Model.
/// </summary>
///
/// <param name="function">The link function to use.</param>
/// <param name="inputs">The number of input variables for the model.</param>
/// <param name="intercept">The starting intercept value. Default is 0.</param>
///
public GeneralizedLinearRegression(ILinkFunction function, int inputs, double intercept)
: this(function, inputs)
{
this.coefficients[0] = intercept;
}
/// <summary>
/// Creates a new Generalized Linear Regression Model.
/// </summary>
///
/// <param name="function">The link function to use.</param>
/// <param name="coefficients">The coefficient vector.</param>
/// <param name="standardErrors">The standard error vector.</param>
///
public GeneralizedLinearRegression(ILinkFunction function,
double[] coefficients, double[] standardErrors)
{
this.linkFunction = function;
this.coefficients = coefficients;
this.standardErrors = standardErrors;
}
public GeneralizedLinearRegression(ILinkFunction function, int inputs, double intercept)
: this(function)
{
this.NumberOfInputs = inputs;
this.Intercept = intercept;
}
