org.apache.commons.math.random
Class RandomDataImpl

java.lang.Object
  org.apache.commons.math.random.RandomDataImpl

All Implemented Interfaces:

java.io.Serializable, RandomData

public class RandomDataImpl
extends java.lang.Object
implements RandomData, java.io.Serializable

Implements the RandomData interface using a RandomGenerator instance to generate non-secure data and a SecureRandom instance to provide data for the nextSecureXxx methods. If no RandomGenerator is provided in the constructor, the default is to use a generator based on Random. To plug in a different implementation, either implement RandomGenerator directly or extend AbstractRandomGenerator.

Supports reseeding the underlying pseudo-random number generator (PRNG). The SecurityProvider and Algorithm used by the SecureRandom instance can also be reset.

For details on the default PRNGs, see Random and SecureRandom.

Usage Notes:

• Instance variables are used to maintain RandomGenerator and SecureRandom instances used in data generation. Therefore, to generate a random sequence of values or strings, you should use just one RandomDataImpl instance repeatedly.
• The "secure" methods are *much* slower. These should be used only when a cryptographically secure random sequence is required. A secure random sequence is a sequence of pseudo-random values which, in addition to being well-dispersed (so no subsequence of values is an any more likely than other subsequence of the the same length), also has the additional property that knowledge of values generated up to any point in the sequence does not make it any easier to predict subsequent values.
• When a new RandomDataImpl is created, the underlying random number generators are not initialized. If you do not explicitly seed the default non-secure generator, it is seeded with the current time in milliseconds on first use. The same holds for the secure generator. If you provide a RandomGenerator to the constructor, however, this generator is not reseeded by the constructor nor is it reseeded on first use.
• The reSeed and reSeedSecure methods delegate to the corresponding methods on the underlying RandomGenerator and SecureRandom instances. Therefore, reSeed(long) fully resets the initial state of the non-secure random number generator (so that reseeding with a specific value always results in the same subsequent random sequence); whereas reSeedSecure(long) does not reinitialize the secure random number generator (so secure sequences started with calls to reseedSecure(long) won't be identical).
• This implementation is not synchronized.

Version:

$Id: RandomDataImpl.java 1182787 2011-10-13 11:20:48Z celestin $

See Also:

Serialized Form

Constructor Summary

RandomDataImpl()
Construct a RandomDataImpl.

RandomDataImpl(RandomGenerator rand)
Construct a RandomDataImpl using the supplied RandomGenerator as the source of (non-secure) random data.

Method Summary

double	nextBeta(double alpha, double beta) Generates a random value from the Beta Distribution.
int	nextBinomial(int numberOfTrials, double probabilityOfSuccess) Generates a random value from the Binomial Distribution.
double	nextCauchy(double median, double scale) Generates a random value from the Cauchy Distribution.
double	nextChiSquare(double df) Generates a random value from the ChiSquare Distribution.
double	nextExponential(double mean) Returns a random value from an Exponential distribution with the given mean.
double	nextF(double numeratorDf, double denominatorDf) Generates a random value from the F Distribution.
double	nextGamma(double shape, double scale) Generates a random value from the Gamma Distribution.
double	nextGaussian(double mu, double sigma) Generate a random value from a Normal (a.k.a.
java.lang.String	nextHexString(int len) Generates a random string of hex characters of length len.
int	nextHypergeometric(int populationSize, int numberOfSuccesses, int sampleSize) Generates a random value from the Hypergeometric Distribution.
int	nextInt(int lower, int upper) Generate a random int value uniformly distributed between lower and upper, inclusive.
double	nextInversionDeviate(ContinuousDistribution distribution) Generate a random deviate from the given distribution using the inversion method.
int	nextInversionDeviate(IntegerDistribution distribution) Generate a random deviate from the given distribution using the inversion method.
long	nextLong(long lower, long upper) Generate a random long value uniformly distributed between lower and upper, inclusive.
int	nextPascal(int r, double p) Generates a random value from the Pascal Distribution.
int[]	nextPermutation(int n, int k) Generates an integer array of length k whose entries are selected randomly, without repetition, from the integers 0 through n-1 (inclusive).
long	nextPoisson(double mean) Generates a random value from the Poisson distribution with the given mean.
java.lang.Object[]	nextSample(java.util.Collection<?> c, int k) Uses a 2-cycle permutation shuffle to generate a random permutation.
java.lang.String	nextSecureHexString(int len) Generates a random string of hex characters from a secure random sequence.
int	nextSecureInt(int lower, int upper) Generate a random int value uniformly distributed between lower and upper, inclusive.
long	nextSecureLong(long lower, long upper) Generate a random long value uniformly distributed between lower and upper, inclusive.
double	nextT(double df) Generates a random value from the T Distribution.
double	nextUniform(double lower, double upper) Generates a uniformly distributed random value from the open interval (lower,upper) (i.e., endpoints excluded).
double	nextWeibull(double shape, double scale) Generates a random value from the Weibull Distribution.
int	nextZipf(int numberOfElements, double exponent) Generates a random value from the Zipf Distribution.
void	reSeed() Reseeds the random number generator with the current time in milliseconds.
void	reSeed(long seed) Reseeds the random number generator with the supplied seed.
void	reSeedSecure() Reseeds the secure random number generator with the current time in milliseconds.
void	reSeedSecure(long seed) Reseeds the secure random number generator with the supplied seed.
void	setSecureAlgorithm(java.lang.String algorithm, java.lang.String provider) Sets the PRNG algorithm for the underlying SecureRandom instance using the Security Provider API.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

RandomDataImpl

public RandomDataImpl()

Construct a RandomDataImpl.

RandomDataImpl

public RandomDataImpl(RandomGenerator rand)

Construct a RandomDataImpl using the supplied RandomGenerator as the source of (non-secure) random data.

Parameters:

rand - the source of (non-secure) random data (may be null, resulting in default JDK-supplied generator)

Since:

1.1

Method Detail

nextHexString

public java.lang.String nextHexString(int len)

Generates a random string of hex characters of length len.

The generated string will be random, but not cryptographically secure. To generate cryptographically secure strings, use nextSecureHexString

Preconditions:

• len > 0 (otherwise an IllegalArgumentException is thrown.)

Algorithm Description: hex strings are generated using a 2-step process.

1. len/2+1 binary bytes are generated using the underlying Random
2. Each binary byte is translated into 2 hex digits

Specified by:

nextHexString in interface RandomData

Parameters:

len - the desired string length.

Returns:

the random string.

Throws:

NotStrictlyPositiveException - if len <= 0.

nextInt

public int nextInt(int lower,
                   int upper)

Generate a random int value uniformly distributed between lower and upper, inclusive.

Specified by:

nextInt in interface RandomData

Parameters:

lower - the lower bound.

upper - the upper bound.

Returns:

the random integer.

Throws:

NumberIsTooLargeException - if lower >= upper.

nextLong

public long nextLong(long lower,
                     long upper)

Generate a random long value uniformly distributed between lower and upper, inclusive.

Specified by:

nextLong in interface RandomData

Parameters:

lower - the lower bound.

upper - the upper bound.

Returns:

the random integer.

Throws:

NumberIsTooLargeException - if lower >= upper.

nextSecureHexString

public java.lang.String nextSecureHexString(int len)

Generates a random string of hex characters from a secure random sequence.

If cryptographic security is not required, use nextHexString().

Preconditions:

• len > 0 (otherwise an IllegalArgumentException is thrown.)

Algorithm Description: hex strings are generated in 40-byte segments using a 3-step process.

1. 20 random bytes are generated using the underlying SecureRandom.
2. SHA-1 hash is applied to yield a 20-byte binary digest.
3. Each byte of the binary digest is converted to 2 hex digits.

Specified by:

nextSecureHexString in interface RandomData

Parameters:

len - the length of the generated string

Returns:

the random string

Throws:

NotStrictlyPositiveException - if len <= 0.

nextSecureInt

public int nextSecureInt(int lower,
                         int upper)

Generate a random int value uniformly distributed between lower and upper, inclusive. This algorithm uses a secure random number generator.

Specified by:

nextSecureInt in interface RandomData

Parameters:

lower - the lower bound.

upper - the upper bound.

Returns:

the random integer.

Throws:

NumberIsTooLargeException - if lower >= upper.

nextSecureLong

public long nextSecureLong(long lower,
                           long upper)

Generate a random long value uniformly distributed between lower and upper, inclusive. This algorithm uses a secure random number generator.

Specified by:

nextSecureLong in interface RandomData

Parameters:

lower - the lower bound.

upper - the upper bound.

Returns:

the random integer.

Throws:

NumberIsTooLargeException - if lower >= upper.

nextPoisson

public long nextPoisson(double mean)

Generates a random value from the Poisson distribution with the given mean.

Definition: Poisson Distribution

Preconditions:

• The specified mean must be positive (otherwise an IllegalArgumentException is thrown.)

Algorithm Description:

• For small means, uses simulation of a Poisson process using Uniform deviates, as described here. The Poisson process (and hence value returned) is bounded by 1000 * mean.
• For large means, uses the rejection algorithm described in
  Devroye, Luc. (1981).The Computer Generation of Poisson Random Variables Computing vol. 26 pp. 197-207.

Specified by:

nextPoisson in interface RandomData

Parameters:

mean - mean of the Poisson distribution.

Returns:

the random Poisson value.

Throws:

NotStrictlyPositiveException - if mean <= 0.

nextGaussian

public double nextGaussian(double mu,
                           double sigma)

Generate a random value from a Normal (a.k.a. Gaussian) distribution with the given mean, mu and the given standard deviation, sigma.

Specified by:

nextGaussian in interface RandomData

Parameters:

mu - the mean of the distribution

sigma - the standard deviation of the distribution

Returns:

the random Normal value

Throws:

NotStrictlyPositiveException - if sigma <= 0.

nextExponential

public double nextExponential(double mean)

Returns a random value from an Exponential distribution with the given mean.

Algorithm Description: Uses the Algorithm SA (Ahrens) from p. 876 in: [1]: Ahrens, J. H. and Dieter, U. (1972). Computer methods for sampling from the exponential and normal distributions. Communications of the ACM, 15, 873-882.

Specified by:

nextExponential in interface RandomData

Parameters:

mean - the mean of the distribution

Returns:

the random Exponential value

Throws:

NotStrictlyPositiveException - if mean <= 0.

nextUniform

public double nextUniform(double lower,
                          double upper)

Generates a uniformly distributed random value from the open interval (lower,upper) (i.e., endpoints excluded).

Definition: Uniform Distribution lower and upper - lower are the location and scale parameters, respectively.

Preconditions:

• lower < upper (otherwise an IllegalArgumentException is thrown.)

Algorithm Description: scales the output of Random.nextDouble(), but rejects 0 values (i.e., will generate another random double if Random.nextDouble() returns 0). This is necessary to provide a symmetric output interval (both endpoints excluded).

Specified by:

nextUniform in interface RandomData

Parameters:

lower - the lower bound.

upper - the upper bound.

Returns:

a uniformly distributed random value from the interval (lower, upper)

Throws:

NumberIsTooLargeException - if lower >= upper.

nextBeta

public double nextBeta(double alpha,
                       double beta)

Generates a random value from the Beta Distribution. This implementation uses inversion to generate random values.

Parameters:

alpha - first distribution shape parameter

beta - second distribution shape parameter

Returns:

random value sampled from the beta(alpha, beta) distribution

Since:

2.2

nextBinomial

public int nextBinomial(int numberOfTrials,
                        double probabilityOfSuccess)

Generates a random value from the Binomial Distribution. This implementation uses inversion to generate random values.

Parameters:

numberOfTrials - number of trials of the Binomial distribution

probabilityOfSuccess - probability of success of the Binomial distribution

Returns:

random value sampled from the Binomial(numberOfTrials, probabilityOfSuccess) distribution

Since:

2.2

nextCauchy

public double nextCauchy(double median,
                         double scale)

Generates a random value from the Cauchy Distribution. This implementation uses inversion to generate random values.

Parameters:

median - the median of the Cauchy distribution

scale - the scale parameter of the Cauchy distribution

Returns:

random value sampled from the Cauchy(median, scale) distribution

Since:

2.2

nextChiSquare

public double nextChiSquare(double df)

Generates a random value from the ChiSquare Distribution. This implementation uses inversion to generate random values.

Parameters:

df - the degrees of freedom of the ChiSquare distribution

Returns:

random value sampled from the ChiSquare(df) distribution

Since:

2.2

nextF

public double nextF(double numeratorDf,
                    double denominatorDf)

Generates a random value from the F Distribution. This implementation uses inversion to generate random values.

Parameters:

numeratorDf - the numerator degrees of freedom of the F distribution

denominatorDf - the denominator degrees of freedom of the F distribution

Returns:

random value sampled from the F(numeratorDf, denominatorDf) distribution

Since:

2.2

nextGamma

public double nextGamma(double shape,
                        double scale)

Generates a random value from the Gamma Distribution.

This implementation uses the following algorithms:

For 0 < shape < 1:
Ahrens, J. H. and Dieter, U., Computer methods for sampling from gamma, beta, Poisson and binomial distributions. Computing, 12, 223-246, 1974.

For shape >= 1:
Marsaglia and Tsang, A Simple Method for Generating Gamma Variables. ACM Transactions on Mathematical Software, Volume 26 Issue 3, September, 2000.

Parameters:

shape - the median of the Gamma distribution

scale - the scale parameter of the Gamma distribution

Returns:

random value sampled from the Gamma(shape, scale) distribution

Since:

2.2

nextHypergeometric

public int nextHypergeometric(int populationSize,
                              int numberOfSuccesses,
                              int sampleSize)

Generates a random value from the Hypergeometric Distribution. This implementation uses inversion to generate random values.

Parameters:

populationSize - the population size of the Hypergeometric distribution

numberOfSuccesses - number of successes in the population of the Hypergeometric distribution

sampleSize - the sample size of the Hypergeometric distribution

Returns:

random value sampled from the Hypergeometric(numberOfSuccesses, sampleSize) distribution

Since:

2.2

nextPascal

public int nextPascal(int r,
                      double p)

Generates a random value from the Pascal Distribution. This implementation uses inversion to generate random values.

Parameters:

r - the number of successes of the Pascal distribution

p - the probability of success of the Pascal distribution

Returns:

random value sampled from the Pascal(r, p) distribution

Since:

2.2

nextT

public double nextT(double df)

Generates a random value from the T Distribution. This implementation uses inversion to generate random values.

Parameters:

df - the degrees of freedom of the T distribution

Returns:

random value from the T(df) distribution

Since:

2.2

nextWeibull

public double nextWeibull(double shape,
                          double scale)

Generates a random value from the Weibull Distribution. This implementation uses inversion to generate random values.

Parameters:

shape - the shape parameter of the Weibull distribution

scale - the scale parameter of the Weibull distribution

Returns:

random value sampled from the Weibull(shape, size) distribution

Since:

2.2

nextZipf

public int nextZipf(int numberOfElements,
                    double exponent)

Generates a random value from the Zipf Distribution. This implementation uses inversion to generate random values.

Parameters:

numberOfElements - the number of elements of the ZipfDistribution

exponent - the exponent of the ZipfDistribution

Returns:

random value sampled from the Zipf(numberOfElements, exponent) distribution

Since:

2.2

reSeed

public void reSeed(long seed)

Reseeds the random number generator with the supplied seed.

Will create and initialize if null.

Parameters:

seed - the seed value to use

reSeedSecure

public void reSeedSecure()

Reseeds the secure random number generator with the current time in milliseconds.

Will create and initialize if null.

reSeedSecure

public void reSeedSecure(long seed)

Reseeds the secure random number generator with the supplied seed.

Will create and initialize if null.

Parameters:

seed - the seed value to use

reSeed

public void reSeed()

Reseeds the random number generator with the current time in milliseconds.

setSecureAlgorithm

public void setSecureAlgorithm(java.lang.String algorithm,
                               java.lang.String provider)
                        throws java.security.NoSuchAlgorithmException,
                               java.security.NoSuchProviderException

Sets the PRNG algorithm for the underlying SecureRandom instance using the Security Provider API. The Security Provider API is defined in Java Cryptography Architecture API Specification & Reference.

USAGE NOTE: This method carries significant overhead and may take several seconds to execute.

Parameters:

algorithm - the name of the PRNG algorithm

provider - the name of the provider

Throws:

java.security.NoSuchAlgorithmException - if the specified algorithm is not available

java.security.NoSuchProviderException - if the specified provider is not installed

nextPermutation

public int[] nextPermutation(int n,
                             int k)

Generates an integer array of length k whose entries are selected randomly, without repetition, from the integers 0 through n-1 (inclusive).

Generated arrays represent permutations of n taken k at a time.

Preconditions:

• k <= n
• n > 0

If the preconditions are not met, an IllegalArgumentException is thrown.

Uses a 2-cycle permutation shuffle. The shuffling process is described here.

Specified by:

nextPermutation in interface RandomData

Parameters:

n - domain of the permutation (must be positive)

k - size of the permutation (must satisfy 0 < k <= n).

Returns:

the random permutation as an int array

Throws:

NumberIsTooLargeException - if k > n.

NotStrictlyPositiveException - if k <= 0.

nextSample

public java.lang.Object[] nextSample(java.util.Collection<?> c,
                                     int k)

Uses a 2-cycle permutation shuffle to generate a random permutation. Algorithm Description: Uses a 2-cycle permutation shuffle to generate a random permutation of c.size() and then returns the elements whose indexes correspond to the elements of the generated permutation. This technique is described, and proven to generate random samples, here

Specified by:

nextSample in interface RandomData

Parameters:

c - Collection to sample from.

k - sample size.

Returns:

the random sample.

Throws:

NumberIsTooLargeException - if k > c.size().

NotStrictlyPositiveException - if k <= 0.

nextInversionDeviate

public double nextInversionDeviate(ContinuousDistribution distribution)

Generate a random deviate from the given distribution using the inversion method.

Parameters:

distribution - Continuous distribution to generate a random value from

Returns:

a random value sampled from the given distribution

Since:

2.2

nextInversionDeviate

public int nextInversionDeviate(IntegerDistribution distribution)

Generate a random deviate from the given distribution using the inversion method.

Parameters:

distribution - Integer distribution to generate a random value from

Returns:

a random value sampled from the given distribution

Since:

2.2