Merge remote-tracking branch 'upstream/master' into windows-test

Author: MarcoGorelli

.github/workflows/arviz_compat.yml

@@ -9,12 +9,12 @@ jobs:
   pytest:
     strategy:
       matrix:
-        python-version: ["3.8"]
         os: [ubuntu-latest, macos-latest]
         floatx: [float64]
         test-subset:
           - pymc3/tests/test_distributions_random.py
           - pymc3/tests/test_sampling.py
+      fail-fast: false
     runs-on: ${{ matrix.os }}
     env:
       TEST_SUBSET: ${{ matrix.test-subset }}
@@ -27,20 +27,28 @@ jobs:
       - name: Cache conda
         uses: actions/cache@v1
         env:
-          # Increase this value to reset cache if environment-dev.yml has not changed
+          # Increase this value to reset cache if environment-dev-py38.yml has not changed
           CACHE_NUMBER: 0
         with:
           path: ~/conda_pkgs_dir
           key: ${{ runner.os }}-conda-${{ env.CACHE_NUMBER }}-${{
-            hashFiles('environment-dev.yml') }}
+            hashFiles('conda-envs/environment-dev-py38.yml') }}
       - uses: conda-incubator/setup-miniconda@v2
         with:
-          activate-environment: testenv
+          activate-environment: pymc3-dev-py38
           channel-priority: strict
-          environment-file: environment-dev.yml
+          environment-file: conda-envs/environment-dev-py38.yml
           use-only-tar-bz2: true # IMPORTANT: This needs to be set for caching to work properly!
-      - run: |
-          conda activate testenv
+      - name: Install pymc3
+        run: |
+          conda activate pymc3-dev-py38
+          pip install -e .
+          python --version
+      - name: Install latest arviz
+        run: |
+          conda activate pymc3-dev-py38
           conda remove arviz -y
           pip install git+git://github.com/arviz-devs/arviz.git
+      - name: Run tests
+        run: |
           python -m pytest -vv --cov=pymc3 --cov-report=xml --cov-report term --durations=50 $TEST_SUBSET

.github/workflows/pytest.yml

@@ -9,7 +9,6 @@ jobs:
   pytest:
     strategy:
       matrix:
-        python-version: ["3.6"]
         os: [ubuntu-18.04]
         floatx: [float32, float64]
         test-subset:
@@ -36,7 +35,6 @@ jobs:
             pymc3/tests/test_dist_math.py
             pymc3/tests/test_distribution_defaults.py
             pymc3/tests/test_distributions_random.py
-            pymc3/tests/test_distributions_timeseries.py
             pymc3/tests/test_parallel_sampling.py
             pymc3/tests/test_random.py
             pymc3/tests/test_shared.py
@@ -47,13 +45,15 @@ jobs:
             pymc3/tests/test_posteriors.py
             pymc3/tests/test_quadpotential.py
           - |
+            pymc3/tests/test_distributions_timeseries.py
             pymc3/tests/test_shape_handling.py
             pymc3/tests/test_updates.py
             pymc3/tests/test_variational_inference.py
           - |
             pymc3/tests/test_distributions.py
             pymc3/tests/test_gp.py
             pymc3/tests/test_sampling.py
+      fail-fast: false
     runs-on: ${{ matrix.os }}
     env:
       TEST_SUBSET: ${{ matrix.test-subset }}
@@ -66,20 +66,26 @@ jobs:
       - name: Cache conda
         uses: actions/cache@v1
         env:
-          # Increase this value to reset cache if environment-dev.yml has not changed
+          # Increase this value to reset cache if environment-dev-py36.yml has not changed
           CACHE_NUMBER: 0
         with:
           path: ~/conda_pkgs_dir
           key: ${{ runner.os }}-conda-${{ env.CACHE_NUMBER }}-${{
-            hashFiles('environment-dev.yml') }}
+            hashFiles('conda-envs/environment-dev-py36.yml') }}
       - uses: conda-incubator/setup-miniconda@v2
         with:
-          activate-environment: testenv
+          activate-environment: pymc3-dev-py36
           channel-priority: strict
-          environment-file: environment-dev.yml
+          environment-file: conda-envs/environment-dev-py36.yml
           use-only-tar-bz2: true # IMPORTANT: This needs to be set for caching to work properly!
-      - run: |
-          conda activate testenv
+      - name: Install-pymc3
+        run: |
+          conda activate pymc3-dev-py36
+          pip install -e .
+          python --version
+      - name: Run tests
+        run: |
+          conda activate pymc3-dev-py36
           python -m pytest -vv --cov=pymc3 --cov-report=xml --cov-report term --durations=50 $TEST_SUBSET
       - name: Upload coverage to Codecov
         uses: codecov/codecov-action@v1

.pre-commit-config.yaml

@@ -41,7 +41,6 @@ repos:
       entry: python scripts/check_toc_is_complete.py
       language: python
       name: Check all notebooks appear in table of contents
-      pass_filenames: false
       types: [jupyter]
     - id: check-no-tests-are-ignored
       entry: python scripts/check_all_tests_are_covered.py

CONTRIBUTING.md

@@ -38,19 +38,27 @@ The preferred workflow for contributing to PyMC3 is to fork the [GitHub reposito
 3. Create a ``feature`` branch to hold your development changes:
 
    ```bash
-   $ git checkout -b my-feature
+   $ git switch -c my-feature
    ```
 
    Always use a ``feature`` branch. It's good practice to never routinely work on the ``master`` branch of any repository.
 
-4. Project requirements are in ``requirements.txt``, and libraries used for development are in ``requirements-dev.txt``.  To set up a development environment, you may (probably in a [virtual environment](https://docs.python-guide.org/dev/virtualenvs/)) run:
+4. Project requirements are in ``requirements.txt``, and libraries used for development are in ``requirements-dev.txt``. The easiest (and recommended) way to set up a development environment is via [miniconda](https://docs.conda.io/en/latest/miniconda.html):
 
    ```bash
-   $ pip install -r requirements.txt
+   $ conda env create -f conda-envs/environment-dev-py36.yml  # or py37 or py38
+   $ conda activate pymc3-dev-py36
+   $ pip install -e .
+   ```
+
+   _Alternatively_ you may (probably in a [virtual environment](https://docs.python-guide.org/dev/virtualenvs/)) run:
+
+   ```bash
+   $ pip install -e .
    $ pip install -r requirements-dev.txt
    ```
 
-Alternatively, there is a script to create a docker environment for development.  See: [Developing in Docker](#Developing-in-Docker).
+   Yet another alternative is to create a docker environment for development. See: [Developing in Docker](#Developing-in-Docker).
 
 5. Develop the feature on your feature branch. Add changed files using ``git add`` and then ``git commit`` files:
 
@@ -103,26 +111,7 @@ tools:
   $ pytest --cov=pymc3 pymc3/tests/<name of test>.py
   ```
 
-* No `pyflakes` warnings, check with:
-
-  ```bash
-  $ pip install pyflakes
-  $ pyflakes path/to/module.py
-  ```
-
-* No PEP8 warnings, check with:
-
-  ```bash
-  $ pip install pycodestyle
-  $ pycodestyle path/to/module.py
-  ```
-
-* AutoPEP8 can help you fix some of the easy redundant errors:
-
-  ```bash
-  $ pip install autopep8
-  $ autopep8 path/to/pep8.py
-  ```
+* No `pre-commit` errors: see the [Python code style](https://github.com/pymc-devs/pymc3/wiki/PyMC3-Python-Code-Style) and [Jupyter Notebook style](https://github.com/pymc-devs/pymc3/wiki/PyMC's-Jupyter-Notebook-Style) page from our Wiki on how to install and run it.
 
 ## Developing in Docker
 

README.rst

@@ -168,6 +168,7 @@ Software using PyMC3
 - `beat <https://github.com/hvasbath/beat>`__: Bayesian Earthquake Analysis Tool.
 - `pymc-learn <https://github.com/pymc-learn/pymc-learn>`__: Custom PyMC models built on top of pymc3_models/scikit-learn API
 - `fenics-pymc3 <https://github.com/IvanYashchuk/fenics-pymc3>`__: Differentiable interface to FEniCS, a library for solving partial differential equations.
+- `cell2location <https://github.com/BayraktarLab/cell2location>`__: Comprehensive mapping of tissue cell architecture via integrated single cell and spatial transcriptomics.
 
 Please contact us if your software is not listed here.
 

RELEASE-NOTES.md

@@ -35,6 +35,7 @@ This new version of `Theano-PyMC` comes with an experimental JAX backend which,
 - Change SMC metropolis kernel to independent metropolis kernel [#4115](https://github.com/pymc-devs/pymc3/pull/4115))
 - Add alternative parametrization to NegativeBinomial distribution in terms of n and p (see [#4126](https://github.com/pymc-devs/pymc3/issues/4126))
 - Added semantically meaningful `str` representations to PyMC3 objects for console, notebook, and GraphViz use (see [#4076](https://github.com/pymc-devs/pymc3/pull/4076), [#4065](https://github.com/pymc-devs/pymc3/pull/4065), [#4159](https://github.com/pymc-devs/pymc3/pull/4159), [#4217](https://github.com/pymc-devs/pymc3/pull/4217), and [#4243](https://github.com/pymc-devs/pymc3/pull/4243)).
+- Add Discrete HyperGeometric Distribution (see [#4249](https://github.com/pymc-devs/pymc3/pull/#4249))
 
 ### Maintenance
 - Switch the dependency of Theano to our own fork, [Theano-PyMC](https://github.com/pymc-devs/Theano-PyMC).
@@ -45,11 +46,14 @@ This new version of `Theano-PyMC` comes with an experimental JAX backend which,
 - Enabled the `Multinomial` distribution to handle batch sizes that have more than 2 dimensions. [#4169](https://github.com/pymc-devs/pymc3/pull/4169)
 - Test model logp before starting any MCMC chains (see [#4116](https://github.com/pymc-devs/pymc3/issues/4116))
 - Fix bug in `model.check_test_point` that caused the `test_point` argument to be ignored. (see [PR #4211](https://github.com/pymc-devs/pymc3/pull/4211#issuecomment-727142721))
+- Refactored MvNormal.random method with better handling of sample, batch and event shapes. [#4207](https://github.com/pymc-devs/pymc3/pull/4207)
+- The `InverseGamma` distribution now implements a `logcdf`. [#3944](https://github.com/pymc-devs/pymc3/pull/3944)
 
 ### Documentation
 - Added a new notebook demonstrating how to incorporate sampling from a conjugate Dirichlet-multinomial posterior density in conjunction with other step methods (see [#4199](https://github.com/pymc-devs/pymc3/pull/4199)).
 - Mentioned the way to do any random walk with `theano.tensor.cumsum()` in `GaussianRandomWalk` docstrings (see [#4048](https://github.com/pymc-devs/pymc3/pull/4048)).
 
+**Release manager** for 3.10.0: Eelke Spaak ([@Spaak](https://github.com/Spaak))
 
 ## PyMC3 3.9.3 (11 August 2020)
 
@@ -69,6 +73,7 @@ This new version of `Theano-PyMC` comes with an experimental JAX backend which,
 
 _NB: The `docs/*` folder is still removed from the tarball due to an upload size limit on PyPi._
 
+**Release manager** for 3.9.3: Kyle Beauchamp ([@kyleabeauchamp](https://github.com/kyleabeauchamp))
 
 ## PyMC3 3.9.2 (24 June 2020)
 
@@ -81,10 +86,14 @@ _NB: The `docs/*` folder is still removed from the tarball due to an upload size
 
 _NB: The `docs/*` folder is still removed from the tarball due to an upload size limit on PyPi._
 
+**Release manager** for 3.9.2: Alex Andorra ([@AlexAndorra](https://github.com/AlexAndorra))
+
 ## PyMC3 3.9.1 (16 June 2020)
 The `v3.9.0` upload to PyPI didn't include a tarball, which is fixed in this release.
 Though we had to temporarily remove the `docs/*` folder from the tarball due to a size limit.
 
+**Release manager** for 3.9.1: Michael Osthege ([@michaelosthege](https://github.com/michaelosthege))
+
 ## PyMC3 3.9.0 (16 June 2020)
 
 ### New features
@@ -130,6 +139,7 @@ Though we had to temporarily remove the `docs/*` folder from the tarball due to
 - Dropped some deprecated kwargs and functions (see [#3906](https://github.com/pymc-devs/pymc3/pull/3906))
 - Dropped the outdated 'nuts' initialization method for `pm.sample` (see [#3863](https://github.com/pymc-devs/pymc3/pull/3863)).
 
+**Release manager** for 3.9.0: Michael Osthege ([@michaelosthege](https://github.com/michaelosthege))
 
 ## PyMC3 3.8 (November 29 2019)
 

environment-dev.yml renamed to conda-envs/environment-dev-py36.yml

@@ -1,36 +1,29 @@
-name: testenv
+name: pymc3-dev-py36
 channels:
   - conda-forge
   - defaults
 dependencies:
-  - python>=3.6
+  - python=3.6
   - arviz>=0.9
   - theano-pymc==1.0.11
   - numpy>=1.13
   - scipy>=0.18
-  - pandas >=0.18
+  - pandas>=0.18
   - patsy>=0.5
   - fastprogress>=0.2
   - h5py>=2.7
   - typing-extensions>=3.7
-  - bokeh>=0.12
-  - coverage>=5.1
   - python-graphviz
   - ipython>=7.16
   - nbsphinx>=0.4
-  - nose>=1.3
-  - nose-parameterized>=0.6
   - numpydoc>=0.9
   - pre-commit>=2.8.0
   - pytest-cov>=2.5
   - pytest>=3.0
   - recommonmark>=0.4
-  - seaborn>=0.8
   - sphinx-autobuild>=0.7
   - sphinx>=1.5
   - watermark
-  - parameterized
-  - ipywidgets
   - dataclasses  # python_version < 3.7
   - contextvars  # python_version < 3.7
   - mkl-service

conda-envs/environment-dev-py37.yml

@@ -0,0 +1,29 @@
+name: pymc3-dev-py37
+channels:
+  - conda-forge
+  - defaults
+dependencies:
+  - python=3.7
+  - arviz>=0.9
+  - theano-pymc==1.0.11
+  - numpy>=1.13
+  - scipy>=0.18
+  - pandas>=0.18
+  - patsy>=0.5
+  - fastprogress>=0.2
+  - h5py>=2.7
+  - typing-extensions>=3.7
+  - python-graphviz
+  - ipython>=7.16
+  - nbsphinx>=0.4
+  - numpydoc>=0.9
+  - pre-commit>=2.8.0
+  - pytest-cov>=2.5
+  - pytest>=3.0
+  - recommonmark>=0.4
+  - sphinx-autobuild>=0.7
+  - sphinx>=1.5
+  - watermark
+  - mkl-service
+  - dill
+  - libblas=*=*mkl

conda-envs/environment-dev-py38.yml

@@ -0,0 +1,29 @@
+name: pymc3-dev-py38
+channels:
+  - conda-forge
+  - defaults
+dependencies:
+  - python=3.8
+  - arviz>=0.9
+  - theano-pymc==1.0.11
+  - numpy>=1.13
+  - scipy>=0.18
+  - pandas>=0.18
+  - patsy>=0.5
+  - fastprogress>=0.2
+  - h5py>=2.7
+  - typing-extensions>=3.7
+  - python-graphviz
+  - ipython>=7.16
+  - nbsphinx>=0.4
+  - numpydoc>=0.9
+  - pre-commit>=2.8.0
+  - pytest-cov>=2.5
+  - pytest>=3.0
+  - recommonmark>=0.4
+  - sphinx-autobuild>=0.7
+  - sphinx>=1.5
+  - watermark
+  - mkl-service
+  - dill
+  - libblas=*=*mkl

docs/source/PyMC3_and_Theano.rst

@@ -11,8 +11,12 @@ build up a Theano function from the space of our parameters to
 their posterior probability density up to a constant factor. We then use
 symbolic manipulations of this function to also get access to its gradient.
 
+Note that the original developers have stopped maintaining Theano, so
+PyMC3 uses `Theano-PyMC <https://github.com/pymc-devs/Theano-PyMC>`_,
+a fork of Theano maintained by the PyMC3 developers.
+
 For a thorough introduction to Theano see the
-`theano docs <http://deeplearning.net/software/theano/introduction.html>`_,
+`theano docs <https://theano-pymc.readthedocs.io/en/latest/>`_,
 but for the most part you don't need detailed knowledge about it as long
 as you are not trying to define new distributions or other extensions
 of PyMC3. But let's look at a simple example to get a rough

docs/source/api/distributions/transforms.rst

@@ -24,7 +24,6 @@ implemented as ``pm.distributions.transforms.``\*X*.
    ordered
    log
    sum_to_1
-   t_stick_breaking
    circular
    CholeskyCovPacked
    Chain
@@ -93,10 +92,6 @@ below.
     :class:`~pymc3.distributions.transforms.Circular` :class:`~pymc3.distributions.transforms.Transform` class
     for use in the ``transform`` argument of a random variable.
 
-
-.. autofunction::  t_stick_breaking
-
-
 Transform Base Classes
 ~~~~~~~~~~~~~~~~~~~~~~
 

docs/source/api/plots.rst

@@ -3,6 +3,7 @@ Plots
 *****
 
 .. currentmodule:: pymc3.plots
+
 Plots are delegated to the
 `ArviZ <https://arviz-devs.github.io/arviz/index.html>`_.
 library, a general purpose library for