This is an automated email from the ASF dual-hosted git repository.
aherbert pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/commons-rng.git
commit a71ba4abc2b514ec0d93ab1f808621a0d5bd5ccf
Author: aherbert <aherb...@apache.org>
AuthorDate: Mon Feb 25 13:11:03 2019 +0000
Document use of null RandomGenerator in the distribution samplers tests.
---
.../distribution/ContinuousSamplersList.java | 78 ++++++++++++----------
.../distribution/DiscreteSamplersList.java | 44 ++++++------
2 files changed, 65 insertions(+), 57 deletions(-)
diff --git
a/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/ContinuousSamplersList.java
b/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/ContinuousSamplersList.java
index 1cf4313..d94b617 100644
---
a/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/ContinuousSamplersList.java
+++
b/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/ContinuousSamplersList.java
@@ -33,176 +33,180 @@ public class ContinuousSamplersList {
static {
try {
- // The commons-math distributions are not used for sampling so use
a null random generator
- org.apache.commons.math3.random.RandomGenerator rng = null;
+ // This test uses reference distributions from commons-math3 to
compute the expected
+ // PMF. These distributions have a dual functionality to compute
the PMF and perform
+ // sampling. When no sampling is needed for the created
distribution, it is advised
+ // to pass null as the random generator via the appropriate
constructors to avoid the
+ // additional initialisation overhead.
+ org.apache.commons.math3.random.RandomGenerator unusedRng = null;
// List of distributions to test.
// Gaussian ("inverse method").
final double meanNormal = -123.45;
final double sigmaNormal = 6.789;
- add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(rng, meanNormal,
sigmaNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(unusedRng, meanNormal,
sigmaNormal),
RandomSource.create(RandomSource.KISS));
// Gaussian (DEPRECATED "Box-Muller").
- add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(rng, meanNormal,
sigmaNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(unusedRng, meanNormal,
sigmaNormal),
new
BoxMullerGaussianSampler(RandomSource.create(RandomSource.MT), meanNormal,
sigmaNormal));
// Gaussian ("Box-Muller").
- add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(rng, meanNormal,
sigmaNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(unusedRng, meanNormal,
sigmaNormal),
new GaussianSampler(new
BoxMullerNormalizedGaussianSampler(RandomSource.create(RandomSource.MT)),
meanNormal, sigmaNormal));
// Gaussian ("Marsaglia").
- add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(rng, meanNormal,
sigmaNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(unusedRng, meanNormal,
sigmaNormal),
new GaussianSampler(new
MarsagliaNormalizedGaussianSampler(RandomSource.create(RandomSource.MT)),
meanNormal, sigmaNormal));
// Gaussian ("Ziggurat").
- add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(rng, meanNormal,
sigmaNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.NormalDistribution(unusedRng, meanNormal,
sigmaNormal),
new GaussianSampler(new
ZigguratNormalizedGaussianSampler(RandomSource.create(RandomSource.MT)),
meanNormal, sigmaNormal));
// Beta ("inverse method").
final double alphaBeta = 4.3;
final double betaBeta = 2.1;
- add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(rng, alphaBeta,
betaBeta),
+ add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(unusedRng, alphaBeta,
betaBeta),
RandomSource.create(RandomSource.ISAAC));
// Beta ("Cheng").
- add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(rng, alphaBeta,
betaBeta),
+ add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(unusedRng, alphaBeta,
betaBeta),
new
ChengBetaSampler(RandomSource.create(RandomSource.MWC_256), alphaBeta,
betaBeta));
- add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(rng, betaBeta,
alphaBeta),
+ add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(unusedRng, betaBeta,
alphaBeta),
new
ChengBetaSampler(RandomSource.create(RandomSource.WELL_19937_A), betaBeta,
alphaBeta));
// Beta ("Cheng", alternate algorithm).
final double alphaBetaAlt = 0.5678;
final double betaBetaAlt = 0.1234;
- add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(rng, alphaBetaAlt,
betaBetaAlt),
+ add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(unusedRng, alphaBetaAlt,
betaBetaAlt),
new
ChengBetaSampler(RandomSource.create(RandomSource.WELL_512_A), alphaBetaAlt,
betaBetaAlt));
- add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(rng, betaBetaAlt,
alphaBetaAlt),
+ add(LIST, new
org.apache.commons.math3.distribution.BetaDistribution(unusedRng, betaBetaAlt,
alphaBetaAlt),
new
ChengBetaSampler(RandomSource.create(RandomSource.WELL_19937_C), betaBetaAlt,
alphaBetaAlt));
// Cauchy ("inverse method").
final double medianCauchy = 0.123;
final double scaleCauchy = 4.5;
- add(LIST, new
org.apache.commons.math3.distribution.CauchyDistribution(rng, medianCauchy,
scaleCauchy),
+ add(LIST, new
org.apache.commons.math3.distribution.CauchyDistribution(unusedRng,
medianCauchy, scaleCauchy),
RandomSource.create(RandomSource.WELL_19937_C));
// Chi-square ("inverse method").
final int dofChi2 = 12;
- add(LIST, new
org.apache.commons.math3.distribution.ChiSquaredDistribution(rng, dofChi2),
+ add(LIST, new
org.apache.commons.math3.distribution.ChiSquaredDistribution(unusedRng,
dofChi2),
RandomSource.create(RandomSource.WELL_19937_A));
// Exponential ("inverse method").
final double meanExp = 3.45;
- add(LIST, new
org.apache.commons.math3.distribution.ExponentialDistribution(rng, meanExp),
+ add(LIST, new
org.apache.commons.math3.distribution.ExponentialDistribution(unusedRng,
meanExp),
RandomSource.create(RandomSource.WELL_44497_A));
// Exponential.
- add(LIST, new
org.apache.commons.math3.distribution.ExponentialDistribution(rng, meanExp),
+ add(LIST, new
org.apache.commons.math3.distribution.ExponentialDistribution(unusedRng,
meanExp),
new
AhrensDieterExponentialSampler(RandomSource.create(RandomSource.MT), meanExp));
// F ("inverse method").
final int numDofF = 4;
final int denomDofF = 7;
- add(LIST, new
org.apache.commons.math3.distribution.FDistribution(rng, numDofF, denomDofF),
+ add(LIST, new
org.apache.commons.math3.distribution.FDistribution(unusedRng, numDofF,
denomDofF),
RandomSource.create(RandomSource.MT_64));
// Gamma ("inverse method").
final double thetaGammaSmallerThanOne = 0.1234;
final double thetaGammaLargerThanOne = 2.345;
final double alphaGamma = 3.456;
- add(LIST, new
org.apache.commons.math3.distribution.GammaDistribution(rng,
thetaGammaLargerThanOne, alphaGamma),
+ add(LIST, new
org.apache.commons.math3.distribution.GammaDistribution(unusedRng,
thetaGammaLargerThanOne, alphaGamma),
RandomSource.create(RandomSource.SPLIT_MIX_64));
// Gamma (theta < 1).
- add(LIST, new
org.apache.commons.math3.distribution.GammaDistribution(rng,
thetaGammaSmallerThanOne, alphaGamma),
+ add(LIST, new
org.apache.commons.math3.distribution.GammaDistribution(unusedRng,
thetaGammaSmallerThanOne, alphaGamma),
new
AhrensDieterMarsagliaTsangGammaSampler(RandomSource.create(RandomSource.XOR_SHIFT_1024_S),
alphaGamma,
thetaGammaSmallerThanOne));
// Gamma (theta > 1).
- add(LIST, new
org.apache.commons.math3.distribution.GammaDistribution(rng,
thetaGammaLargerThanOne, alphaGamma),
+ add(LIST, new
org.apache.commons.math3.distribution.GammaDistribution(unusedRng,
thetaGammaLargerThanOne, alphaGamma),
new
AhrensDieterMarsagliaTsangGammaSampler(RandomSource.create(RandomSource.WELL_44497_B),
alphaGamma,
thetaGammaLargerThanOne));
// Gumbel ("inverse method").
final double muGumbel = -4.56;
final double betaGumbel = 0.123;
- add(LIST, new
org.apache.commons.math3.distribution.GumbelDistribution(rng, muGumbel,
betaGumbel),
+ add(LIST, new
org.apache.commons.math3.distribution.GumbelDistribution(unusedRng, muGumbel,
betaGumbel),
RandomSource.create(RandomSource.WELL_1024_A));
// Laplace ("inverse method").
final double muLaplace = 12.3;
final double betaLaplace = 5.6;
- add(LIST, new
org.apache.commons.math3.distribution.LaplaceDistribution(rng, muLaplace,
betaLaplace),
+ add(LIST, new
org.apache.commons.math3.distribution.LaplaceDistribution(unusedRng, muLaplace,
betaLaplace),
RandomSource.create(RandomSource.MWC_256));
// Levy ("inverse method").
final double muLevy = -1.098;
final double cLevy = 0.76;
- add(LIST, new
org.apache.commons.math3.distribution.LevyDistribution(rng, muLevy, cLevy),
+ add(LIST, new
org.apache.commons.math3.distribution.LevyDistribution(unusedRng, muLevy,
cLevy),
RandomSource.create(RandomSource.TWO_CMRES));
// Log normal ("inverse method").
final double scaleLogNormal = 2.345;
final double shapeLogNormal = 0.1234;
- add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(rng,
scaleLogNormal, shapeLogNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(unusedRng,
scaleLogNormal, shapeLogNormal),
RandomSource.create(RandomSource.KISS));
// Log-normal (DEPRECATED "Box-Muller").
- add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(rng,
scaleLogNormal, shapeLogNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(unusedRng,
scaleLogNormal, shapeLogNormal),
new
BoxMullerLogNormalSampler(RandomSource.create(RandomSource.XOR_SHIFT_1024_S),
scaleLogNormal, shapeLogNormal));
// Log-normal ("Box-Muller").
- add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(rng,
scaleLogNormal, shapeLogNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(unusedRng,
scaleLogNormal, shapeLogNormal),
new LogNormalSampler(new
BoxMullerNormalizedGaussianSampler(RandomSource.create(RandomSource.XOR_SHIFT_1024_S)),
scaleLogNormal, shapeLogNormal));
// Log-normal ("Marsaglia").
- add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(rng,
scaleLogNormal, shapeLogNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(unusedRng,
scaleLogNormal, shapeLogNormal),
new LogNormalSampler(new
MarsagliaNormalizedGaussianSampler(RandomSource.create(RandomSource.MT_64)),
scaleLogNormal, shapeLogNormal));
// Log-normal ("Ziggurat").
- add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(rng,
scaleLogNormal, shapeLogNormal),
+ add(LIST, new
org.apache.commons.math3.distribution.LogNormalDistribution(unusedRng,
scaleLogNormal, shapeLogNormal),
new LogNormalSampler(new
ZigguratNormalizedGaussianSampler(RandomSource.create(RandomSource.MWC_256)),
scaleLogNormal, shapeLogNormal));
// Logistic ("inverse method").
final double muLogistic = -123.456;
final double sLogistic = 7.89;
- add(LIST, new
org.apache.commons.math3.distribution.LogisticDistribution(rng, muLogistic,
sLogistic),
+ add(LIST, new
org.apache.commons.math3.distribution.LogisticDistribution(unusedRng,
muLogistic, sLogistic),
RandomSource.create(RandomSource.TWO_CMRES_SELECT, null, 2,
6));
// Nakagami ("inverse method").
final double muNakagami = 78.9;
final double omegaNakagami = 23.4;
final double inverseAbsoluteAccuracyNakagami =
org.apache.commons.math3.distribution.NakagamiDistribution.DEFAULT_INVERSE_ABSOLUTE_ACCURACY;
- add(LIST, new
org.apache.commons.math3.distribution.NakagamiDistribution(rng, muNakagami,
omegaNakagami, inverseAbsoluteAccuracyNakagami),
+ add(LIST, new
org.apache.commons.math3.distribution.NakagamiDistribution(unusedRng,
muNakagami, omegaNakagami, inverseAbsoluteAccuracyNakagami),
RandomSource.create(RandomSource.TWO_CMRES_SELECT, null, 5,
3));
// Pareto ("inverse method").
final double scalePareto = 23.45;
final double shapePareto = 0.1234;
- add(LIST, new
org.apache.commons.math3.distribution.ParetoDistribution(rng, scalePareto,
shapePareto),
+ add(LIST, new
org.apache.commons.math3.distribution.ParetoDistribution(unusedRng,
scalePareto, shapePareto),
RandomSource.create(RandomSource.TWO_CMRES_SELECT, null, 9,
11));
// Pareto.
- add(LIST, new
org.apache.commons.math3.distribution.ParetoDistribution(rng, scalePareto,
shapePareto),
+ add(LIST, new
org.apache.commons.math3.distribution.ParetoDistribution(unusedRng,
scalePareto, shapePareto),
new
InverseTransformParetoSampler(RandomSource.create(RandomSource.XOR_SHIFT_1024_S),
scalePareto, shapePareto));
// T ("inverse method").
final double dofT = 0.76543;
- add(LIST, new
org.apache.commons.math3.distribution.TDistribution(rng, dofT),
+ add(LIST, new
org.apache.commons.math3.distribution.TDistribution(unusedRng, dofT),
RandomSource.create(RandomSource.ISAAC));
// Triangular ("inverse method").
final double aTriangle = -0.76543;
final double cTriangle = -0.65432;
final double bTriangle = -0.54321;
- add(LIST, new
org.apache.commons.math3.distribution.TriangularDistribution(rng, aTriangle,
cTriangle, bTriangle),
+ add(LIST, new
org.apache.commons.math3.distribution.TriangularDistribution(unusedRng,
aTriangle, cTriangle, bTriangle),
RandomSource.create(RandomSource.MT));
// Uniform ("inverse method").
final double loUniform = -1.098;
final double hiUniform = 0.76;
- add(LIST, new
org.apache.commons.math3.distribution.UniformRealDistribution(rng, loUniform,
hiUniform),
+ add(LIST, new
org.apache.commons.math3.distribution.UniformRealDistribution(unusedRng,
loUniform, hiUniform),
RandomSource.create(RandomSource.TWO_CMRES));
// Uniform.
- add(LIST, new
org.apache.commons.math3.distribution.UniformRealDistribution(rng, loUniform,
hiUniform),
+ add(LIST, new
org.apache.commons.math3.distribution.UniformRealDistribution(unusedRng,
loUniform, hiUniform),
new
ContinuousUniformSampler(RandomSource.create(RandomSource.MT_64), loUniform,
hiUniform));
// Weibull ("inverse method").
final double alphaWeibull = 678.9;
final double betaWeibull = 98.76;
- add(LIST, new
org.apache.commons.math3.distribution.WeibullDistribution(rng, alphaWeibull,
betaWeibull),
+ add(LIST, new
org.apache.commons.math3.distribution.WeibullDistribution(unusedRng,
alphaWeibull, betaWeibull),
RandomSource.create(RandomSource.WELL_44497_B));
} catch (Exception e) {
System.err.println("Unexpected exception while creating the list
of samplers: " + e);
diff --git
a/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/DiscreteSamplersList.java
b/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/DiscreteSamplersList.java
index b9e1bc0..5ccc99a 100644
---
a/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/DiscreteSamplersList.java
+++
b/commons-rng-sampling/src/test/java/org/apache/commons/rng/sampling/distribution/DiscreteSamplersList.java
@@ -35,25 +35,29 @@ public class DiscreteSamplersList {
static {
try {
- // The commons-math distributions are not used for sampling so use
a null random generator
- org.apache.commons.math3.random.RandomGenerator rng = null;
+ // This test uses reference distributions from commons-math3 to
compute the expected
+ // PMF. These distributions have a dual functionality to compute
the PMF and perform
+ // sampling. When no sampling is needed for the created
distribution, it is advised
+ // to pass null as the random generator via the appropriate
constructors to avoid the
+ // additional initialisation overhead.
+ org.apache.commons.math3.random.RandomGenerator unusedRng = null;
// List of distributions to test.
// Binomial ("inverse method").
final int trialsBinomial = 20;
final double probSuccessBinomial = 0.67;
- add(LIST, new
org.apache.commons.math3.distribution.BinomialDistribution(rng, trialsBinomial,
probSuccessBinomial),
+ add(LIST, new
org.apache.commons.math3.distribution.BinomialDistribution(unusedRng,
trialsBinomial, probSuccessBinomial),
MathArrays.sequence(8, 9, 1),
RandomSource.create(RandomSource.KISS));
// Geometric ("inverse method").
final double probSuccessGeometric = 0.21;
- add(LIST, new
org.apache.commons.math3.distribution.GeometricDistribution(rng,
probSuccessGeometric),
+ add(LIST, new
org.apache.commons.math3.distribution.GeometricDistribution(unusedRng,
probSuccessGeometric),
MathArrays.sequence(10, 0, 1),
RandomSource.create(RandomSource.ISAAC));
// Geometric.
- add(LIST, new
org.apache.commons.math3.distribution.GeometricDistribution(rng,
probSuccessGeometric),
+ add(LIST, new
org.apache.commons.math3.distribution.GeometricDistribution(unusedRng,
probSuccessGeometric),
MathArrays.sequence(10, 0, 1),
new
GeometricSampler(RandomSource.create(RandomSource.XOR_SHIFT_1024_S),
probSuccessGeometric));
@@ -61,48 +65,48 @@ public class DiscreteSamplersList {
final int popSizeHyper = 34;
final int numSuccessesHyper = 11;
final int sampleSizeHyper = 12;
- add(LIST, new
org.apache.commons.math3.distribution.HypergeometricDistribution(rng,
popSizeHyper, numSuccessesHyper, sampleSizeHyper),
+ add(LIST, new
org.apache.commons.math3.distribution.HypergeometricDistribution(unusedRng,
popSizeHyper, numSuccessesHyper, sampleSizeHyper),
MathArrays.sequence(10, 0, 1),
RandomSource.create(RandomSource.MT));
// Pascal ("inverse method").
final int numSuccessesPascal = 6;
final double probSuccessPascal = 0.2;
- add(LIST, new
org.apache.commons.math3.distribution.PascalDistribution(rng,
numSuccessesPascal, probSuccessPascal),
+ add(LIST, new
org.apache.commons.math3.distribution.PascalDistribution(unusedRng,
numSuccessesPascal, probSuccessPascal),
MathArrays.sequence(18, 1, 1),
RandomSource.create(RandomSource.TWO_CMRES));
// Uniform ("inverse method").
final int loUniform = -3;
final int hiUniform = 4;
- add(LIST, new
org.apache.commons.math3.distribution.UniformIntegerDistribution(rng,
loUniform, hiUniform),
+ add(LIST, new
org.apache.commons.math3.distribution.UniformIntegerDistribution(unusedRng,
loUniform, hiUniform),
MathArrays.sequence(8, -3, 1),
RandomSource.create(RandomSource.SPLIT_MIX_64));
// Uniform.
- add(LIST, new
org.apache.commons.math3.distribution.UniformIntegerDistribution(rng,
loUniform, hiUniform),
+ add(LIST, new
org.apache.commons.math3.distribution.UniformIntegerDistribution(unusedRng,
loUniform, hiUniform),
MathArrays.sequence(8, -3, 1),
new
DiscreteUniformSampler(RandomSource.create(RandomSource.MT_64), loUniform,
hiUniform));
// Uniform (large range).
final int halfMax = Integer.MAX_VALUE / 2;
final int hiLargeUniform = halfMax + 10;
final int loLargeUniform = -hiLargeUniform;
- add(LIST, new
org.apache.commons.math3.distribution.UniformIntegerDistribution(rng,
loLargeUniform, hiLargeUniform),
+ add(LIST, new
org.apache.commons.math3.distribution.UniformIntegerDistribution(unusedRng,
loLargeUniform, hiLargeUniform),
MathArrays.sequence(20, -halfMax, halfMax / 10),
new
DiscreteUniformSampler(RandomSource.create(RandomSource.WELL_1024_A),
loLargeUniform, hiLargeUniform));
// Zipf ("inverse method").
final int numElementsZipf = 5;
final double exponentZipf = 2.345;
- add(LIST, new
org.apache.commons.math3.distribution.ZipfDistribution(rng, numElementsZipf,
exponentZipf),
+ add(LIST, new
org.apache.commons.math3.distribution.ZipfDistribution(unusedRng,
numElementsZipf, exponentZipf),
MathArrays.sequence(5, 1, 1),
RandomSource.create(RandomSource.XOR_SHIFT_1024_S));
// Zipf.
- add(LIST, new
org.apache.commons.math3.distribution.ZipfDistribution(rng, numElementsZipf,
exponentZipf),
+ add(LIST, new
org.apache.commons.math3.distribution.ZipfDistribution(unusedRng,
numElementsZipf, exponentZipf),
MathArrays.sequence(5, 1, 1),
new
RejectionInversionZipfSampler(RandomSource.create(RandomSource.WELL_19937_C),
numElementsZipf, exponentZipf));
// Zipf (exponent close to 1).
final double exponentCloseToOneZipf = 1 - 1e-10;
- add(LIST, new
org.apache.commons.math3.distribution.ZipfDistribution(rng, numElementsZipf,
exponentCloseToOneZipf),
+ add(LIST, new
org.apache.commons.math3.distribution.ZipfDistribution(unusedRng,
numElementsZipf, exponentCloseToOneZipf),
MathArrays.sequence(5, 1, 1),
new
RejectionInversionZipfSampler(RandomSource.create(RandomSource.WELL_19937_C),
numElementsZipf, exponentCloseToOneZipf));
@@ -110,33 +114,33 @@ public class DiscreteSamplersList {
final double epsilonPoisson =
org.apache.commons.math3.distribution.PoissonDistribution.DEFAULT_EPSILON;
final int maxIterationsPoisson =
org.apache.commons.math3.distribution.PoissonDistribution.DEFAULT_MAX_ITERATIONS;
final double meanPoisson = 3.21;
- add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(rng, meanPoisson,
epsilonPoisson, maxIterationsPoisson),
+ add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(unusedRng,
meanPoisson, epsilonPoisson, maxIterationsPoisson),
MathArrays.sequence(10, 0, 1),
RandomSource.create(RandomSource.MWC_256));
// Poisson.
- add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(rng, meanPoisson,
epsilonPoisson, maxIterationsPoisson),
+ add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(unusedRng,
meanPoisson, epsilonPoisson, maxIterationsPoisson),
MathArrays.sequence(10, 0, 1),
new PoissonSampler(RandomSource.create(RandomSource.KISS),
meanPoisson));
// Dedicated small mean poisson sampler
- add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(rng, meanPoisson,
epsilonPoisson, maxIterationsPoisson),
+ add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(unusedRng,
meanPoisson, epsilonPoisson, maxIterationsPoisson),
MathArrays.sequence(10, 0, 1),
new
SmallMeanPoissonSampler(RandomSource.create(RandomSource.KISS), meanPoisson));
// Poisson (40 < mean < 80).
final double largeMeanPoisson = 67.89;
- add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(rng,
largeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
+ add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(unusedRng,
largeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
MathArrays.sequence(50, (int) (largeMeanPoisson - 25), 1),
new
PoissonSampler(RandomSource.create(RandomSource.SPLIT_MIX_64),
largeMeanPoisson));
// Dedicated large mean poisson sampler
- add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(rng,
largeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
+ add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(unusedRng,
largeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
MathArrays.sequence(50, (int) (largeMeanPoisson - 25), 1),
new
LargeMeanPoissonSampler(RandomSource.create(RandomSource.SPLIT_MIX_64),
largeMeanPoisson));
// Poisson (mean >> 40).
final double veryLargeMeanPoisson = 543.21;
- add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(rng,
veryLargeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
+ add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(unusedRng,
veryLargeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
MathArrays.sequence(100, (int) (veryLargeMeanPoisson - 50), 1),
new
PoissonSampler(RandomSource.create(RandomSource.SPLIT_MIX_64),
veryLargeMeanPoisson));
// Dedicated large mean poisson sampler
- add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(rng,
veryLargeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
+ add(LIST, new
org.apache.commons.math3.distribution.PoissonDistribution(unusedRng,
veryLargeMeanPoisson, epsilonPoisson, maxIterationsPoisson),
MathArrays.sequence(100, (int) (veryLargeMeanPoisson - 50), 1),
new
LargeMeanPoissonSampler(RandomSource.create(RandomSource.SPLIT_MIX_64),
veryLargeMeanPoisson));
} catch (Exception e) {
