diff options
| author | Robert Kern <robert.kern@gmail.com> | 2021-05-05 10:27:24 -0400 |
|---|---|---|
| committer | Robert Kern <robert.kern@gmail.com> | 2021-05-05 10:27:24 -0400 |
| commit | d97c634a04e539de2b66ccd0fbd526fea310571a (patch) | |
| tree | 2b808c22d39b13c8b54234d4399ec1d3639c3be8 /numpy | |
| parent | d82ef187439081088ab606da712628959e711a3c (diff) | |
| download | numpy-d97c634a04e539de2b66ccd0fbd526fea310571a.tar.gz | |
DOC: more completely describe the implemented variant.
Diffstat (limited to 'numpy')
| -rw-r--r-- | numpy/random/_pcg64.pyx | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/numpy/random/_pcg64.pyx b/numpy/random/_pcg64.pyx index aaa79b304..8a00dc265 100644 --- a/numpy/random/_pcg64.pyx +++ b/numpy/random/_pcg64.pyx @@ -302,7 +302,10 @@ cdef class PCG64DXSM(BitGenerator): PCG-64 DXSM is a 128-bit implementation of O'Neill's permutation congruential generator ([1]_, [2]_). PCG-64 DXSM has a period of :math:`2^{128}` and supports advancing an arbitrary number of steps as well as :math:`2^{127}` streams. - The specific member of the PCG family that we use is PCG DXSM 128/64. + The specific member of the PCG family that we use is PCG CM DXSM 128/64. It + differs from ``PCG64`` in that it uses the stronger DXSM output function, + a 64-bit "cheap multiplier" in the LCG, and outputs from the state before + advancing it rather than advance-then-output. ``PCG64DXSM`` provides a capsule containing function pointers that produce doubles, and unsigned 32 and 64- bit integers. These are not |