Constructor Detail

Method Detail

• aph_simplify

   final static Pair<DoubleArray, Matrix> aph_simplify(DoubleArray a1, Matrix T1, DoubleArray a2, Matrix T2, Double p1, Double p2, ConvolutionPattern pattern)

  Simplifies/combines two APH distributions using a specified pattern.

  Parameters:

  a1 - Initial probability vector of first distribution

  T1 - Rate matrix of first distribution

  a2 - Initial probability vector of second distribution

  T2 - Rate matrix of second distribution

  p1 - Branch probability for first distribution (for BRANCH pattern)

  p2 - Branch probability for second distribution (for BRANCH pattern)

  pattern - Convolution pattern to use

  Returns:

  Pair of (combined alpha, combined T)

• aph_convpara

   final static Pair<DoubleArray, Matrix> aph_convpara(List<Pair<DoubleArray, Matrix>> distributions)

  Performs convolution on parallel structure with any number of elements.

  Parameters:

  distributions - List of pairs (alpha, T) for each distribution

  Returns:

  Combined (alpha, T)

• aph_convseq

   final static Pair<DoubleArray, Matrix> aph_convseq(List<Pair<DoubleArray, Matrix>> distributions)

  Performs convolution on sequential structure with any number of elements.

  Parameters:

  distributions - List of pairs (alpha, T) for each distribution

  Returns:

  Combined (alpha, T)

• aph_rand

   final static MatrixCell aph_rand(Integer K)

  Generates a random APH (acyclic phase-type) distribution as a MAP.

  Parameters:

  K - Order of the APH distribution (default: 2)

  Returns:

  MAP representation {D0, D1}

• aph_fit

   final static Pair<MatrixCell, Boolean> aph_fit(Double e1, Double e2, Double e3, Integer nmax)

  Fits an APH distribution to match first three moments.

  Implementation based on: A.Bobbio, A.Horvath, M.Telek, "Matching three moments with minimal acyclic phase type distributions", Stochastic Models 21:303-326, 2005.

  Parameters:

  e1 - First moment (mean)

  e2 - Second moment

  e3 - Third moment

  nmax - Maximum order to try (default: 10)

  Returns:

  Pair of (fitted MAP, isExact flag)

• ph2hyper

   final static Pair<DoubleArray, DoubleArray> ph2hyper(MatrixCell PH)

  Converts a hyper-exponential PH distribution to its rate/probability form.

  Parameters:

  PH - Phase-type distribution as {D0, D1}

  Returns:

  Pair of (rates, probabilities)

• hyper_rand

   final static DoubleArray hyper_rand(DoubleArray rates, DoubleArray probs, Integer nSamples)

  Generates random samples from a hyper-exponential distribution.

  Parameters:

  rates - Exponential rates

  probs - Selection probabilities

  nSamples - Number of samples

  Returns:

  Array of samples