Merge pull request #317 from bashtage/remove-generator

Remove generator

MANIFEST.in

@@ -1,4 +1,5 @@
 recursive-exclude randomgen  *.c
+include randomgen/py.typed
 include randomgen/_version.py
 include requirements.txt
 include README.md

README.md

@@ -37,9 +37,13 @@ or the [stable documentation](https://bashtage.github.io/randomgen/devel/change-
 
 # WARNINGS
 
-## Changes in v1.19
+## Changes in v1.24
 
-``Generator`` and ``RandomState`` have been officially deprecated, and will
+``Generator`` and ``RandomState`` have been **removed**.
+
+## Changes from 1.18 to 1.19
+
+``Generator`` and ``RandomState`` have been officially deprecated in 1.19, and will
 warn with a ``FutureWarning`` about their removal. They will also receive virtually
 no maintenance. It is now time to move to NumPy's ``np.random.Generator`` which has
 features not in ``randomstate.Generator`` and is maintained more actively.
@@ -54,8 +58,7 @@ to ``randomstate.ExtendedGenerator``:
 There are no plans to remove any of the bit generators, e.g., ``AESCounter``,
 ``ThreeFry``, or ``PCG64``. 
 
-## Changes in v1.18
-
+### Changes from 1.16 to 1.18
 There are many changes between v1.16.x and v1.18.x. These reflect API
 decision taken in conjunction with NumPy in preparation of the core
 of `randomgen` being used as the preferred random number generator in
@@ -66,16 +69,6 @@ bit generators (or `BigGenerator`s).
 
 ## Future Plans
 
-A substantial portion of randomgen has been merged into NumPy. Revamping NumPy's random
-number generation was always the goal of this project (and its predecessor
-[NextGen NumPy RandomState](https://github.com/bashtage/ng-numpy-randomstate>)),
-and so it has succeeded.
-
-While I have no immediate plans to remove anything, after a 1.19 release I will:
-
-* Remove `Generator` and `RandomState`. These duplicate NumPy and will diverge over time.
-  The versions in NumPy are authoritative. **Deprecated**
-* Preserve novel methods of `Generator` in a new class, `ExtendedGenerator`. **Done**
 * Add some distributions that are not supported in NumPy. _Ongoing_
 * Add any interesting bit generators I come across. _Recent additions include the DXSM and CM-DXSM variants of PCG64 and the LXM generator._
 
@@ -102,7 +95,6 @@ The RNGs include:
 * Chaotic PRNGS: Small-Fast Chaotic (`SFC64`) and Jenkin's Small-Fast (`JSF`).
 
 
-
 ## Status
 
 * Builds and passes all tests on:
@@ -114,8 +106,8 @@ The RNGs include:
 
 ## Version
 
-The package version matches the latest version of NumPy where
-`Generator(MT19937())` passes all NumPy test.
+The package version matches the latest version of NumPy when the package
+is released.
 
 ## Documentation
 
@@ -128,8 +120,8 @@ the latest commit (unreleased) is available under
 ## Requirements
 Building requires:
 
-* Python (3.6, 3.7, 3.8, 3.9)
-* NumPy (1.14, 1.15, 1.16, 1.17, 1.18, 1.19)
+* Python (3.6, 3.7, 3.8, 3.9, 3.10)
+* NumPy (1.17+)
 * Cython (0.29+)
 * tempita (0.5+), if not provided by Cython
 

appveyor.yml

@@ -28,7 +28,3 @@ build_script:
 
 test_script:
   - pytest randomgen
-
-on_success:
-  - cd %GIT_DIR%\
-  - python benchmark.py

azure-pipelines.yml

@@ -21,14 +21,14 @@ jobs:
 - template: ci/azure/azure_template_posix.yml
   parameters:
     name: macOS
-    vmImage: macOS-10.15
+    vmImage: macOS-latest
 
 - template: ci/azure/azure_template_posix.yml
   parameters:
     name: Linux
-    vmImage: ubuntu-20.04
+    vmImage: ubuntu-latest
 
 - template: ci/azure/azure_template_windows.yml
   parameters:
     name: Windows
-    vmImage: windows-2019
+    vmImage: windows-latest

benchmark.py

@@ -15,7 +15,8 @@
     rg = numpy.random.RandomState()
     rg.random_sample()
 else:
-    from randomgen import Generator, {bitgen}
+    from randomgen import {bitgen}
+    from numpy.random import Generator
     rg = Generator({bitgen}())
     rg.random()
 """

ci/azure/azure_template_posix.yml

@@ -25,7 +25,7 @@ jobs:
         python38_legacy:
           python.version: '3.8'
           coverage: true
-          NUMPY: 1.16.6
+          NUMPY: 1.17.0
         python39_latest:
           python.version: '3.9'
         python310_latest:

ci/azure/install-posix.sh

@@ -13,13 +13,13 @@ else
 fi
 
 # Not all available in conda
-python -m pip install setuptools "setuptools_scm[toml]<7" "oldest-supported-numpy" wheel pip black==22.3.0 isort flake8 --upgrade
+python -m pip install setuptools "setuptools_scm[toml]<7" "oldest-supported-numpy" wheel pip black==22.6.0 isort flake8 --upgrade
 
 EXTRA="pytest pytest-xdist coverage pytest-cov"
 
-if [[ -n ${NUMPY} ]]; then CMD="$CMD==${NUMPY}"; fi;
+if [[ -n ${NUMPY} ]]; then CMD="$CMD~=${NUMPY}"; fi;
 CMD="$CMD cython"
-if [[ -n ${CYTHON} ]]; then CMD="$CMD==${CYTHON}"; fi;
+if [[ -n ${CYTHON} ]]; then CMD="$CMD~=${CYTHON}"; fi;
 CMD="$CMD pandas"
 CMD="$CMD $EXTRA"
 if [[ ${USE_CONDA} == "true" ]]; then CMD="$CMD numba"; fi;

doc/requirements.txt

@@ -1,7 +1,7 @@
 numba>=0.49
 nbsphinx
 sphinx-material
-sphinx>=3
+sphinx>=5
 ipython>=6
 numpydoc
 pygments>=2.7.4 # not directly required, pinned by Snyk to avoid a vulnerability

doc/source/bit_generators/index.rst

@@ -2,7 +2,7 @@ Bit Generators
 --------------
 
 The random values produced by :class:`numpy.random.Generator`
-(and also  :class:`~randomgen.generator.Generator`)
+(and also  ``Generator``)
 are produced by a bit generator.  These bit generators do not directly provide
 random numbers and only contain methods used for seeding, getting or
 setting the state, jumping or advancing the state, and for accessing

doc/source/change-log.rst

@@ -9,11 +9,13 @@ Change Log
 
       <p class="admonition-title"> Deprecated </p>
 
-  :class:`~randomgen.generator.Generator` and :class:`~randomgen.mtrand.RandomState`
-  are **DEPRECATED**. You should be using :class:`numpy.random.Generator` or
-  :class:`numpy.random.RandomState` which are better maintained. These will be
-  maintained until after NumPy 1.21 (or 2 releases after NumPy 1.19) for users who
-  cannot update NumPy.
+  ``Generator`` and ``RandomState`` have been **REMOVED** in 1.23.
+  You should be using :class:`numpy.random.Generator` or
+  :class:`numpy.random.RandomState` which are maintained. 
+
+v1.23.0
+=======
+- Removed ``Generator`` and ``RandomState``.
 
 v1.20.2
 =======
@@ -44,7 +46,7 @@ v1.20.2
 
 v1.20.1
 =======
-- Fixed a bug that affects :func:`~randomgen.generator.Generator.standard_gamma` when
+- Fixed a bug that affects ``standard_gamma`` when
   used with ``out`` and a Fortran contiguous array.
 - Added :func:`~randomgen.generator.ExtendedGenerator.multivariate_complex_normal`.
 - Added :func:`~randomgen.generator.ExtendedGenerator.standard_wishart` and
@@ -105,20 +107,20 @@ v1.19.0
 - Added support for broadcasting inputs in :class:`randomgen.generator.ExtendedGenerator.multivariate_normal`.
 - Added support for the `++` variant of :class:`randomgen.xoroshiro128.Xoroshiro128`.
 - Fixed a bug the produced incorrect results in :func:`~randomgen.mt19937.MT19937.jumped`.
-- Fixed multiple bugs in :class:`~randomgen.generator.Generator` that were fixed in :class:`numpy.random.Generator`.
+- Fixed multiple bugs in ``Generator`` that were fixed in :class:`numpy.random.Generator`.
 
 v1.18.0
 =======
-- :meth:`~randomgen.generator.Generator.choice` pulled in upstream performance improvement that
+- ``choice`` pulled in upstream performance improvement that
   use a hash set when choosing without replacement and without user-provided probabilities.
 - Added support for :class:`~randomgen.seed_sequence.SeedSequence` (and NumPy's ``SeedSequence``).
 - Fixed a bug that affected both :class:`~randomgen.generator.Generator.randint`
-  in :class:`~randomgen.generator.Generator` and :meth:`~randomgen.mtrand.RandomState.randint`
-  in  :class:`~randomgen.mtrand.RandomState` when ``high=2**32``.  This value is inbounds for
+  in ``Generator`` and ``randint``
+  in  ``RandomState`` when ``high=2**32``.  This value is inbounds for
   a 32-bit unsigned closed interval generator, and so  should have been redirected to
   a 32-bit generator. It  was erroneously sent to the 64-bit path. The random values produced
   are fully random but inefficient. This fix breaks the stream in :class:`~randomgen.generator.Generator
-  is the value for ``high`` is used. The fix restores :class:`~randomgen.mtrand.RandomState` to
+  is the value for ``high`` is used. The fix restores ``RandomState`` to
   NumPy 1.16 compatibility.
   only affects the output if ``dtype`` is ``'int64'``
 - This release brings many breaking changes.  Most of these have been
@@ -127,20 +129,16 @@ v1.18.0
   going into NumPy.
 - Two changes that are more abrupt are:
 
-  * The ``.generator`` method of the bit generators raise
-    ``NotImplementedError``
+  * The ``.generator`` method of the bit generators raise ``NotImplementedError``
   * The internal structures that is used in C have been renamed.
     The main rename is ``brng_t`` to ``bitgen_t``
 
 - The other key changes are:
 
-  * Rename ``RandomGenerator`` to :class:`~randomgen.generator.Generator`.
-  * Rename :meth:`~randomgen.generator.Generator.randint` to
-    :meth:`~randomgen.generator.Generator.integers`.
-  * Rename :meth:`~randomgen.generator.Generator.random_integers` to
-    :meth:`~randomgen.generator.Generator.integers`.
-  * Rename :meth:`~randomgen.generator.Generator.random_sample`
-    to :meth:`~randomgen.generator.Generator.random`.
+  * Rename ``RandomGenerator`` to ``Generator``.
+  * Rename ``randint`` to ``integers``.
+  * Rename ``random_integers`` to ``integers``.
+  * Rename ``random_sample`` to ``random``.
   * Change ``jump`` which operated in-place to
     :meth:`~randomgen.xoshiro256.Xoshiro256.jumped` which
     returns a new ``BitGenerator``.
@@ -163,40 +161,34 @@ v1.16.6
 - Improved the performance of :class:`~randomgen.pcg64.PCG64` on Windows.
 - Improved performance of :func:`~randomgen.dsfmt.DSFMT.jump` and
   :func:`~randomgen.dsfmt.DSFMT.jumped`.
-- Improves backward compatibility of :class:`~randomgen.mtrand.RandomState`
+- Improves backward compatibility of ``RandomState``
 
 
 v1.16.5
 =======
-- Fixed bugs in :func:`~randomgen.mtrand.RandomState.laplace`,
-  :func:`~randomgen.mtrand.RandomState.gumbel`,
-  :func:`~randomgen.mtrand.RandomState.logseries`,
-  :func:`~randomgen.mtrand.RandomState.normal`,
-  :func:`~randomgen.mtrand.RandomState.standard_normal`,
-  :func:`~randomgen.mtrand.RandomState.standard_exponential`,
-  :func:`~randomgen.mtrand.RandomState.exponential`, and
-  :func:`~randomgen.mtrand.RandomState.logistic` that could result in ``nan``
-  values in rare circumstances (about 1 in :math:`10^{53}` draws).
-- Added keyword ``closed`` to :func:`~randomgen.generator.Generator.randint`
+- Fixed bugs in ``laplace``, ``gumbel``, ``logseries``, ``normal``,
+  ``standard_normal``, ``standard_exponential``, ``exponential``, and ``logistic``
+  that could result in ``nan`` values in rare circumstances (about 1 in :math:`10^{53}` draws).
+- Added keyword ``closed`` to ``randint``
   which changes sampling from the half-open interval ``[low, high)`` to the closed
   interval ``[low, high]``.
-- Fixed a bug in :func:`~randomgen.mtrand.RandomState.random_integers` that
+- Fixed a bug in ``random_integers`` that
   could lead to valid values being treated as invalid.
 
 v1.16.4
 =======
-- Add a fast path for broadcasting :func:`~randomgen.generator.Generator.randint`
+- Add a fast path for broadcasting ``randint``
   when using ``uint64`` or ``int64``.
 - Refactor PCG64 so that it does not rely on Cython conditional compilation.
-- Add :func:`~randomgen.generator.Generator.brng` to access the basic RNG.
-- Allow multidimensional arrays in :func:`~randomgen.generator.Generator.choice`.
-- Speed-up :func:`~randomgen.generator.Generator.choice` when not replacing.
+- Add ``brng`` to access the basic RNG.
+- Allow multidimensional arrays in ``choice``.
+- Speed-up ``choice`` when not replacing.
   The gains can be very large (1000x or more) when the input array is large but
   the sample size is small.
-- Add parameter checks in :func:`~randomgen.generator.Generator.multinomial`.
-- Fix an edge-case bug in :func:`~randomgen.generator.Generator.zipf`.
-- Allow 0 for sample in :func:`~randomgen.generator.Generator.hypergeometric`.
-- Add broadcasting to :func:`~randomgen.generator.Generator.multinomial` (see
+- Add parameter checks in ``multinomial``.
+- Fix an edge-case bug in ``zipf``.
+- Allow 0 for sample in ``hypergeometric``.
+- Add broadcasting to ``multinomial`` (see
   `NumPy issue 9710 <https://github.com/numpy/numpy/pull/9710>`_)
 
 v1.16.3
@@ -210,13 +202,10 @@ v1.16.2
   into NumPy, including removing:
 
   * ``random_raw``, which have been moved to the individual bit generators
-  * ``random_uintegers``, which can be replaced with
-    :func:`~randomgen.generator.Generator.randint`.
+  * ``random_uintegers``, which can be replaced with ``randint``.
 
-- Added :class:`~randomgen.mtrand.RandomState` as a clone of NumPy's
-  RandomState.
-- Removed :class:`~randomgen.legacy.LegacyGenerator` since this is no
-  longer needed
+- Added ``RandomState`` as a clone of NumPy's RandomState.
+- Removed ``LegacyGenerator`` since this is no longer needed
 - Fixed many small bugs, including in cffi and ctype interfaces
 
 v1.16.1