Merge pull request #232 from pymc-labs/pymc5-migration

PyMC 5 migration

Author: twiecki

.github/workflows/release.yml

@@ -17,7 +17,7 @@ jobs:
           fetch-depth: 0
       - uses: actions/setup-python@v4
         with:
-          python-version: 3.8
+          python-version: '3.10'
       - name: Build the sdist and the wheel
         run: |
           pip install build
@@ -40,7 +40,7 @@ jobs:
           echo "Checking import and version number (on release)"
           venv-bdist/bin/python -c "import homepy; assert homepy.__version__ == '${{  github.ref_name }}' if '${{ github.ref_type }}' == 'tag' else homepy.__version__; print(homepy.__version__)"
           cd ..
-      - uses: actions/upload-artifact@v3
+      - uses: actions/upload-artifact@v4
         with:
           name: artifact
           path: dist/*
@@ -63,7 +63,7 @@ jobs:
           CACHE_NUMBER: 0
         with:
           path: ~/conda_pkgs_dir
-          key: ${{ runner.os }}-py3.8-conda-${{ env.CACHE_NUMBER }}-${{
+          key: ${{ runner.os }}-py3.10-conda-${{ env.CACHE_NUMBER }}-${{
             hashFiles('envs/environment-test.yml') }}
       - uses: conda-incubator/setup-miniconda@v2
         with:
@@ -73,13 +73,15 @@ jobs:
           activate-environment: homepy
           channel-priority: strict
           environment-file: envs/environment-test.yml
-          python-version: '3.8'
+          python-version: '3.10'
           use-mamba: true
           use-only-tar-bz2: true # IMPORTANT: This needs to be set for caching to work properly!
       - name: Install-package
         run: |
           pip install -e .[dev]
           python --version
+          # Verify PyMC v5 is installed
+          python -c "import pymc; print(f'PyMC version: {pymc.__version__}'); assert pymc.__version__.startswith('5'), 'PyMC v5 required'"
       - name: Run tests
         run: |
           conda activate homepy
@@ -102,7 +104,7 @@ jobs:
           repository_url: https://test.pypi.org/legacy/
       - uses: actions/setup-python@v4
         with:
-          python-version: 3.8
+          python-version: '3.10'
       - name: Test pip install from test.pypi
         run: |
           # Give time to test.pypi to update its index. If we don't wait,

.github/workflows/tests.yml

@@ -43,7 +43,7 @@ jobs:
     - uses: actions/checkout@v3
     - uses: actions/setup-python@v4
       with:
-        python-version: '3.8'
+        python-version: '3.10'
     - uses: pre-commit/action@v3.0.0
   tests:
     needs:
@@ -57,32 +57,19 @@ jobs:
     strategy:
       fail-fast: true
     steps:
-      - uses: actions/checkout@v2
-      - name: Cache conda
-        uses: actions/cache@v3
-        env:
-          # Increase this value to reset cache if environment-test.yml has not changed
-          CACHE_NUMBER: 0
-        with:
-          path: ~/conda_pkgs_dir
-          key: ${{ runner.os }}-py3.8-conda-${{ env.CACHE_NUMBER }}-${{
-            hashFiles('envs/environment-test.yml') }}
-      - uses: conda-incubator/setup-miniconda@v2
+      - uses: actions/checkout@v3
+      - uses: mamba-org/setup-micromamba@v2
         with:
-          miniforge-variant: Mambaforge
-          miniforge-version: latest
-          mamba-version: "*"
-          activate-environment: homepy
-          channel-priority: strict
           environment-file: envs/environment-test.yml
-          python-version: '3.8'
-          use-mamba: true
-          use-only-tar-bz2: true # IMPORTANT: This needs to be set for caching to work properly!
+          environment-name: homepy-test
+          init-shell: bash
+          cache-environment: true
       - name: Install-package
         run: |
           pip install -e .[dev]
           python --version
+          # Verify PyMC v5 is installed
+          python -c "import pymc; print(f'PyMC version: {pymc.__version__}'); assert pymc.__version__.startswith('5'), 'PyMC v5 required'"
       - name: Run tests
         run: |
-          conda activate homepy
-          pytest
+          python -m pytest -vv --cov=homepy --cov-report=xml --no-cov-on-fail --cov-report term-missing

.pre-commit-config.yaml

@@ -1,7 +1,7 @@
 exclude: ^(docs/logos|homepy/tests|)/
 repos:
 - repo: https://github.com/pre-commit/pre-commit-hooks
-  rev: v4.0.1
+  rev: v5.0.0
   hooks:
     -   id: check-merge-conflict
     -   id: check-toml
@@ -13,46 +13,31 @@ repos:
     -   id: requirements-txt-fixer
         exclude: ^requirements-dev\.txt$
     -   id: trailing-whitespace
-- repo: https://github.com/PyCQA/isort
-  rev: 5.12.0
-  hooks:
-    - id: isort
-      name: isort
-- repo: https://github.com/asottile/pyupgrade
-  rev: v2.29.1
-  hooks:
-    - id: pyupgrade
-      args: [--py37-plus]
-- repo: https://github.com/psf/black
-  rev: 22.3.0
-  hooks:
-    - id: black
-- repo: https://github.com/PyCQA/pylint
-  rev: v2.12.1
-  hooks:
-    - id: pylint
-      args: [--rcfile=.pylintrc]
-      files: ^homepy/
 - repo: https://github.com/MarcoGorelli/madforhooks
-  rev: 0.2.1
+  rev: 0.4.1
   hooks:
     - id: no-print-statements
       files: ^homepy/
     - id: conda-env-sorter
       files: ^conda-envs/environment-dev-py.+\.yml$
-- repo: https://github.com/nbQA-dev/nbQA
-  rev: 1.6.1
-  hooks:
-    - id: nbqa-pyupgrade
-      additional_dependencies: [pyupgrade==2.29.1]
-      args: ["--py37-plus"]
-    - id: nbqa-isort
-      additional_dependencies: [isort==5.12.0]
-    - id: nbqa-black
-      additional_dependencies: [black==22.3.0]
-    - id: nbqa-pylint
-      args: [--rcfile=.pylintrc]
-      additional_dependencies: [pylint==2.12.1]
+- repo: https://github.com/astral-sh/ruff-pre-commit
+  rev: v0.12.7
+  hooks:
+  - id: ruff
+    types_or:
+    - python
+    - pyi
+    - jupyter
+    args:
+    - --fix
+    - --exit-non-zero-on-fix
+    exclude: ^(docs)
+  - id: ruff-format
+    types_or:
+    - python
+    - pyi
+    - jupyter
+    exclude: ^(docs)
 - repo: local
   hooks:
     - id: check-no-tests-are-ignored
@@ -68,3 +53,11 @@ repos:
       files: ^conda-envs/environment-dev-py.+.yml$
       language: python
       name: Generate pip dependency from conda
+- repo: https://github.com/lucianopaz/head_of_apache
+  rev: "0.1.1"
+  hooks:
+    - id: head_of_apache
+      args:
+        - --author
+        - PyMC Labs
+        - --last-year-present

docs/02_modeling.ipynb

@@ -221,7 +221,7 @@
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
-       "<p>9000 rows × 6 columns</p>\n",
+       "<p>9000 rows \u00d7 6 columns</p>\n",
        "</div>"
       ],
       "text/plain": [
@@ -1075,7 +1075,7 @@
     "\n",
     "The `n` parameter is equal to the concentration parameter from our observational distribution. In our case, since we assumed that `n` followed an exponential distribution with unit length scale, it's impossible that it's value were too big.\n",
     "\n",
-    "The `p` parameter is linked to the negative binomial's mean. Under the parametrization that numpy works with, the expected value of the negative binomial is `n * (1 - p) / p`. With some math, we can see that small `p` values imply extremely large expected observations! This means that our prior for the model made it possible to have extremely large means after the link function! Let's see how we can fix that."
+    "The `p` parameter is linked to the negative binomial's mean. Under the parametrization that numpy works with, the expected value of the negative binomial is `n * (1 - p) / p`. With some math, we can see that small `p` values imply extremely large expected observations! This means that our prior for the model made it possible to have extremely large means after the link function! Let's see how we can fix thpt."
    ]
   },
   {
@@ -4857,7 +4857,7 @@
    ],
    "source": [
     "import graphviz\n",
-    "from aesara.graph.basic import walk\n",
+    "from pytensor.graph.basic import walk\n",
     "\n",
     "tracked = [\n",
     "    n\n",

docs/04_customizing.ipynb

@@ -61,7 +61,7 @@
     "import numpy as np\n",
     "import pandas as pd\n",
     "import pymc as pm\n",
-    "from aesara import tensor as at\n",
+    "from pytensor import tensor as at\n",
     "\n",
     "import homepy"
    ]
@@ -795,7 +795,7 @@
     "        coefficients = pm.Normal(name, mu=mu, sigma=sigma, shape=len(regressor_columns))\n",
     "\n",
     "        regressors = pm.ConstantData(\"The regressors\", data[regressor_columns].values)\n",
-    "        return at.dot(regressors, coefficients)"
+    "        return pt.dot(regressors, coefficients)"
    ]
   },
   {
@@ -2033,7 +2033,7 @@
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
-       "<p>9000 rows × 6 columns</p>\n",
+       "<p>9000 rows \u00d7 6 columns</p>\n",
        "</div>"
       ],
       "text/plain": [
@@ -2735,7 +2735,7 @@
     "            d.node.dim_name: inds[d.observation_indexes]\n",
     "            for d in hierarchy.levels[hierarchy.depth_to_level[hierarchy.depth]]\n",
     "        }\n",
-    "        return at.concatenate(\n",
+    "        return pt.concatenate(\n",
     "            self.hierarchy.walk_hierarchy(\n",
     "                root_callback=self.root_callback,\n",
     "                intermediate_callback=self.intermediate_callback,\n",
@@ -2764,7 +2764,7 @@
     "        latent = pm.gp.Latent(cov_func=cov)\n",
     "        return pm.Deterministic(\n",
     "            f\"gp_{dim_name}\",\n",
-    "            at.stack(\n",
+    "            pt.stack(\n",
     "                [\n",
     "                    parent_gp + latent.prior(f\"gp_{dim_name}_{i}\", X)\n",
     "                    for i in hierarchy.coords[dim_name]\n",