Package etomica.util.random

Class RandomNumberGeneratorUnix

java.lang.Object

etomica.util.random.RandomNumberGeneratorUnix

All Implemented Interfaces:

IRandom

public class RandomNumberGeneratorUnix
extends java.lang.Object
implements IRandom

Random number generator that takes random bits from files. This can be used with the unix special files /dev/random or /dev/urandom.

Due to file access, this is likely to be slower than other RNG, but could be used to provide an initial seed.

Field Summary

Fields

Modifier and Type	Field	Description
protected byte[]	buf
protected java.io.FileInputStream	fReader
protected boolean	hasNextGaussian
protected double	nextGaussian

Constructor Summary

Constructors

Constructor	Description
RandomNumberGeneratorUnix()	Create an RNG using /dev/urandom as a source of random bits.
RandomNumberGeneratorUnix(java.lang.String randFile)	Create an RNG using the given file as a source of random bits.

Method Summary

Modifier and Type	Method	Description
void	dispose()	Cleanup, close the file.
protected void	fill(int offset, int numBytes)	Fills the byte buffer with random bytes up to numBytes, starting at offset.
static int[]	getRandSeedArray()	Convenience method that returns an array of integers, useful as seeds to another RNG.
static void	main(java.lang.String[] args)
double	nextDouble()	Returns a random double, uniformly distributed between 0 and 1.
double	nextFixedDouble()	Returns a pseudorandom double, uniformly distributed between 0.0 (inclusive) and 1.0 (exclusive).
float	nextFloat()	Returns a random float, uniformly distributed between 0 and 1.
double	nextGaussian()	Returns a double taken from a Gaussian distribution with 0 mean and variance of 1.
int	nextInt()	Returns a random int (32-bit integer).
int	nextInt(int max)	Returns a random int ranging from 0 to max-1.
long	nextLong()	Returns a random long (64-bit integer).
long	nextLong(long max)	Returns a random long ranging from 0 to max-1.

Methods inherited from class java.lang.Object

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

Field Detail

fReader

protected final java.io.FileInputStream fReader

buf

protected final byte[] buf

hasNextGaussian

protected boolean hasNextGaussian

nextGaussian

protected double nextGaussian

Constructor Detail

RandomNumberGeneratorUnix

public RandomNumberGeneratorUnix()

Create an RNG using /dev/urandom as a source of random bits.

RandomNumberGeneratorUnix

public RandomNumberGeneratorUnix(java.lang.String randFile)

Create an RNG using the given file as a source of random bits.

/dev/urandom as a randFile will not block /dev/random as a randFile may block if the entropy pool is depleted

Method Detail

getRandSeedArray

public static int[] getRandSeedArray()

Convenience method that returns an array of integers, useful as seeds to another RNG. If an instance of this class can be instantiated, it is used to return 4 integers. Otherwise, System.nanotime is split into 2 integers.

main

public static void main(java.lang.String[] args)

dispose

public void dispose()

Cleanup, close the file.

fill

protected void fill(int offset,
                    int numBytes)

Fills the byte buffer with random bytes up to numBytes, starting at offset.

nextLong

public long nextLong()

Returns a random long (64-bit integer).

nextInt

public int nextInt()

Returns a random int (32-bit integer).

nextInt

public int nextInt(int max)

Returns a random int ranging from 0 to max-1.

Specified by:

nextInt in interface IRandom

nextLong

public long nextLong(long max)

Returns a random long ranging from 0 to max-1.

nextFloat

public float nextFloat()

Returns a random float, uniformly distributed between 0 and 1. Neither 0 nor 1 will ever be returned.

nextDouble

public double nextDouble()

Returns a random double, uniformly distributed between 0 and 1. Neither 0 nor 1 will ever be returned.

Specified by:

nextDouble in interface IRandom

nextFixedDouble

public double nextFixedDouble()

Description copied from interface: IRandom

Returns a pseudorandom double, uniformly distributed between 0.0 (inclusive) and 1.0 (exclusive). For computational speed, this implementation returns a fixed-point random number (at least in binary), meaning that it will be a random integer times 2^-53.

Specified by:

nextFixedDouble in interface IRandom

nextGaussian

public double nextGaussian()

Returns a double taken from a Gaussian distribution with 0 mean and variance of 1.

Specified by:

nextGaussian in interface IRandom