RandomGenerator#

TheGenerator provides access toa wide range of distributions, and served as a replacement forRandomState. The main difference betweenthe two is thatGenerator relies on an additional BitGenerator tomanage state and generate the random bits, which are then transformed intorandom values from useful distributions. The default BitGenerator used byGenerator isPCG64. The BitGeneratorcan be changed by passing an instantized BitGenerator toGenerator.

numpy.random.default_rng(seed=None)#

Construct a new Generator with the default BitGenerator (PCG64).

Parameters:

seed{None, int, array_like[ints], SeedSequence, BitGenerator, Generator, RandomState}, optional

A seed to initialize theBitGenerator. If None, then fresh,unpredictable entropy will be pulled from the OS. If anint orarray_like[ints] is passed, then all values must be non-negative and will bepassed toSeedSequence to derive the initialBitGenerator state. One may alsopass in aSeedSequence instance.Additionally, when passed aBitGenerator, it will be wrapped byGenerator. If passed aGenerator, it will be returned unaltered.When passed a legacyRandomState instance it will be coerced to aGenerator.

Returns:

Generator

The initialized generator object.

Notes

Ifseed is not aBitGenerator or aGenerator, a newBitGeneratoris instantiated. This function does not manage a default global instance.

SeeSeeding and entropy for more information about seeding.

Examples

Try it in your browser!

default_rng is the recommended constructor for the random number classGenerator. Here are several ways we can construct a random number generator usingdefault_rng and theGenerator class.

Here we usedefault_rng to generate a random float:

>>>importnumpyasnp>>>rng=np.random.default_rng(12345)>>>print(rng)Generator(PCG64)>>>rfloat=rng.random()>>>rfloat0.22733602246716966>>>type(rfloat)<class 'float'>

Here we usedefault_rng to generate 3 random integers between 0 (inclusive) and 10 (exclusive):

>>>importnumpyasnp>>>rng=np.random.default_rng(12345)>>>rints=rng.integers(low=0,high=10,size=3)>>>rintsarray([6, 2, 7])>>>type(rints[0])<class 'numpy.int64'>

Here we specify a seed so that we have reproducible results:

>>>importnumpyasnp>>>rng=np.random.default_rng(seed=42)>>>print(rng)Generator(PCG64)>>>arr1=rng.random((3,3))>>>arr1array([[0.77395605, 0.43887844, 0.85859792],       [0.69736803, 0.09417735, 0.97562235],       [0.7611397 , 0.78606431, 0.12811363]])

If we exit and restart our Python interpreter, we’ll see that wegenerate the same random numbers again:

>>>importnumpyasnp>>>rng=np.random.default_rng(seed=42)>>>arr2=rng.random((3,3))>>>arr2array([[0.77395605, 0.43887844, 0.85859792],       [0.69736803, 0.09417735, 0.97562235],       [0.7611397 , 0.78606431, 0.12811363]])

Go BackOpen In Tab

classnumpy.random.Generator(bit_generator)#

Container for the BitGenerators.

Generator exposes a number of methods for generating randomnumbers drawn from a variety of probability distributions. In addition tothe distribution-specific arguments, each method takes a keyword argumentsize that defaults toNone. Ifsize isNone, then a singlevalue is generated and returned. Ifsize is an integer, then a 1-Darray filled with generated values is returned. Ifsize is a tuple,then an array with that shape is filled and returned.

The functionnumpy.random.default_rng will instantiateaGenerator with numpy’s defaultBitGenerator.

No Compatibility Guarantee

Generator does not provide a version compatibility guarantee. Inparticular, as better algorithms evolve the bit stream may change.

Parameters:

bit_generatorBitGenerator

BitGenerator to use as the core generator.

See also

default_rng

Recommended constructor forGenerator.

Notes

The Python stdlib modulerandom containspseudo-random number generator with a number of methods that are similarto the ones available inGenerator.It uses Mersenne Twister, and this bit generator canbe accessed usingMT19937.Generator, besides beingNumPy-aware, has the advantage that it provides a much larger numberof probability distributions to choose from.

Examples

Try it in your browser!

>>>fromnumpy.randomimportGenerator,PCG64>>>rng=Generator(PCG64())>>>rng.standard_normal()-0.203  # random

Go BackOpen In Tab

>>>importnumpyasnp>>>rng=np.random.default_rng()>>>x=np.arange(0,15).reshape(3,5)>>>xarray([[ 0,  1,  2,  3,  4],       [ 5,  6,  7,  8,  9],       [10, 11, 12, 13, 14]])>>>y=rng.permuted(x,axis=1,out=x)>>>xarray([[ 1,  0,  2,  4,  3],  # random       [ 6,  7,  8,  9,  5],       [10, 14, 11, 13, 12]])

>>>yisxTrue

>>>importnumpyasnp>>>rng=np.random.default_rng()>>>x=np.arange(0,15).reshape(3,5)>>>xarray([[ 0,  1,  2,  3,  4],       [ 5,  6,  7,  8,  9],       [10, 11, 12, 13, 14]])>>>rng.permutation(x,axis=1)array([[ 1,  3,  2,  0,  4],  # random       [ 6,  8,  7,  5,  9],       [11, 13, 12, 10, 14]])

>>>importnumpyasnp>>>rng=np.random.default_rng()>>>rng.permuted(x,axis=1)array([[ 1,  0,  2,  4,  3],  # random       [ 5,  7,  6,  9,  8],       [10, 14, 12, 13, 11]])

>>>importnumpyasnp>>>rng=np.random.default_rng()>>>a=['A','B','C','D','E']>>>rng.shuffle(a)# shuffle the list in-place>>>a['B', 'D', 'A', 'E', 'C']  # random