What’s New or Different

Differences from NumPy (1.17+)

• UserBitGenerator allows bit generators to be written in Python (slow, suitable for experiments and testing) or numba (fast, similar speed to compiled C). See the demonstration notebook for examples.

• PCG64 supports additional variants of PCG64, including the PCG4 2.0 variant (“cm-dxsm”).

• SFC64 supports optional Weyl sequence increments other than 1 which is the fixed increment in numpy.random.SFC64.

• random_entropy() provides access to the system source of randomness that is used in cryptographic applications (e.g., /dev/urandom on Unix).

• Support broadcasting when producing multivariate Gaussian values (multivariate_normal())

• Simulate from the complex normal distribution (complex_normal())

• Direct access to unsigned integers is provided by (uintegers())

• A wider range of bit generators:

  • Chaotic mappings

    • JSF (32 and 64-bit variants)

    • SFC64

  • Cryptographic Cipher-based:

    • AESCounter

    • ChaCha

    • HC128

    • Philox (limited version in NumPy)

    • SPECK128

    • ThreeFry

  • Hardware-based:

    • RDRAND

  • Mersenne Twisters

    • DSFMT

    • MT64

    • MT19937 (in NumPy)

    • SFMT

  • Permuted Congruential Generators

    • PCG32

    • PCG64 (limited version in NumPy)

    • LCG128Mix (limited version in NumPy)

  • Shift/rotate based:

    • LXM

    • Xoroshiro128

    • Xorshift1024

    • Xoshiro256

    • Xoshiro512

Deprecated

Generator is deprecated. You should be using numpy.random.Generator.

• randomgen’s Generator continues to expose legacy methods random_sample, randint, random_integers, rand, randn, and tomaxint. Note: These should not be used, and their modern replacements are preferred:

  • random_sample, rand` → ``random

  • random_integers, randint → integers

  • randn → standard_normal

  • tomaxint → integers with dtype set to int

• randomgen’s bit generators remain seedable and the convenience function seed` is exposed as part of``Generator. Additionally, the convenience property state is available to get or set the state of the underlying bit generator.

• numpy.random.Generator.multivariate_hypergeometric() was added after Generator was merged into NumPy and will not be ported over.

• numpy.random.Generator.shuffle() and numpy.random.Generator.permutation() support axis keyword to operator along an axis other than 0.

• integers supports the keyword argument use_masked to switch between masked generation of bounded integers and Lemire’s superior method.

Differences from NumPy before 1.17

• The normal, exponential and gamma generators use 256-step Ziggurat methods which are 2-10 times faster than NumPy’s default implementation in standard_normal, standard_exponential or standard_gamma.

• The Box-Muller used to produce NumPy’s normals is no longer available.

• All bit generators functions to produce doubles, uint64s and uint32s via CTypes (ctypes()) and CFFI (cffi()). This allows the bit generators to be used in numba or in other low-level applications

• The bit generators can be used in downstream projects via Cython.

• Optional dtype argument that accepts np.float32 or np.float64 to produce either single or double prevision uniform random variables for select core distributions

  • Uniforms (random and rand)

  • Normals (standard_normal and randn)

  • Standard Gammas (standard_gamma)

  • Standard Exponentials (standard_exponential)

• Optional out argument that allows existing arrays to be filled for select core distributions

  • Uniforms (random)

  • Normals (standard_normal)

  • Standard Gammas (standard_gamma)

  • Standard Exponentials (standard_exponential)

  This allows multithreading to fill large arrays in chunks using suitable PRNGs in parallel.

• integers supports broadcasting inputs.

• integers supports drawing from open (default, [low, high)) or closed ([low, high]) intervals using the keyword argument endpoint. Closed intervals are simpler to use when the distribution may include the maximum value of a given integer type.

• The closed interval is particularly helpful when using arrays since it avoids object-dtype arrays when sampling from the full range.

• Support for Lemire’s method of generating uniform integers on an arbitrary interval by setting use_masked=True in (integers).

• multinomial supports multidimensional values of n

• choice is much faster when sampling small amounts from large arrays

• choice supports the axis keyword to work with multidimensional arrays.

• For changes since the previous release, see the Change Log