Package net.jqwik.api

Interface RandomDistribution

@API(status=MAINTAINED,
     since="1.3.0")
public interface RandomDistribution

Determines how generated numerical values are generated and distributed across the allowed range and a center withing this range.

Since all random numeric value generation is going back to BigDecimal generation this interfaces uses only values of type BigDecimal.

The generation of an arbitrary's edge cases is not influenced by the distribution.

See Also:

BigIntegerArbitrary, IntegerArbitrary, LongArbitrary, ShortArbitrary, ByteArbitrary, BigDecimalArbitrary, DoubleArbitrary, FloatArbitrary

Nested Class Summary

Nested Classes

Modifier and Type	Interface	Description
static class	RandomDistribution.RandomDistributionFacade
static interface	RandomDistribution.RandomNumericGenerator	Generator for BigInteger values which are behind all generated numeric values in jqwik.

Method Summary

Modifier and Type	Method	Description
static RandomDistribution	biased()	A distribution that generates values closer to the center of a numerical range with a higher probability.
RandomDistribution.RandomNumericGenerator	createGenerator(int genSize, java.math.BigInteger min, java.math.BigInteger max, java.math.BigInteger center)	Return a generator that will generate value with the desired distribution
static RandomDistribution	gaussian()	A gaussian distribution with borderSigma of 3, i.e. approximately 99.7% of values are within the borders.
static RandomDistribution	gaussian(double borderSigma)	A distribution that generates values with (potentially asymmetric) gaussian distribution the mean of which is the specified center and the probability at the borders is approximately borderSigma times standard deviation.
static RandomDistribution	uniform()	A distribution that generates values across the allowed range with a uniform probability distribution.

Method Detail

biased

static RandomDistribution biased()

A distribution that generates values closer to the center of a numerical range with a higher probability. The bigger the range the stronger the bias.

Returns:

a random distribution instance

uniform

static RandomDistribution uniform()

A distribution that generates values across the allowed range with a uniform probability distribution.

Returns:

a random distribution instance

gaussian

static RandomDistribution gaussian(double borderSigma)

A distribution that generates values with (potentially asymmetric) gaussian distribution the mean of which is the specified center and the probability at the borders is approximately borderSigma times standard deviation.

Gaussian generation is approximately 10 times slower than biased() or uniform() generation. But still, except in rare cases this will not make a noticeable difference in the runtime of your properties.

Parameters:

borderSigma - The approximate factor of standard deviation at the border(s)

Returns:

a random distribution instance

gaussian

static RandomDistribution gaussian()

A gaussian distribution with borderSigma of 3, i.e. approximately 99.7% of values are within the borders.

Returns:

a random distribution instance

See Also:

gaussian(double)

createGenerator

@API(status=INTERNAL)
RandomDistribution.RandomNumericGenerator createGenerator(int genSize,
                                                          java.math.BigInteger min,
                                                          java.math.BigInteger max,
                                                          java.math.BigInteger center)

Return a generator that will generate value with the desired distribution

Parameters:

genSize - The approximate number of values to generate. Can be influenced by callers.

min - The minimum allowed value (included)

max - The maximum allowed value (included)

center - The center for the distribution. Must be within min and max.

Returns:

generator for randomly generated BigInteger values