DRAFT: Python implementation of cobyla

.github/workflows/test_python.yml

@@ -0,0 +1,68 @@
+name: Test Python
+
+on:
+  # Trigger the workflow on push or pull request
+  #push:
+  pull_request:  # DANGEROUS! MUST be disabled for self-hosted runners!
+  # Trigger the workflow by cron. The default time zone of GitHub Actions is UTC.
+  schedule:
+    - cron: '0 16 4-31/4 * *'
+  # Trigger the workflow manually
+  workflow_dispatch:
+
+
+jobs:
+
+  test:
+    name: Run Python tests
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest, windows-latest, macos-latest]
+        solver: [cobyla]
+        testdim: [small]
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v3.5.3
+        with:
+          submodules: recursive
+        #   ssh-key: ${{ secrets.SSH_PRIVATE_KEY_ACT }} # This forces checkout to use SSH, not HTTPS
+        # As of 230425, checkout with ssh fails frequently on Windows runners.
+
+      # TODO: See if this is necessary for Python
+      # - name: Make tools such as grep, make, and git available on Windows
+      #   if: runner.os == 'Windows'
+      #   run: $env:Path += ";C:\Program Files\Git\usr\bin;C:\Program Files\Git\bin;C:\ProgramData\Chocolatey\bin"
+
+      # - name: Miscellaneous setup
+      #   shell: bash  # Important; otherwise, the following statements do not work on Windows.
+      #   run: bash .github/scripts/misc_setup
+      - name: Setup Python
+        uses: actions/setup-python@v3
+        with:
+          python-version: "3.11"
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install numpy pytest
+
+      - name: Conduct the test
+        # shell: bash  # Important; otherwise, `<` will not work on Windows.
+        run: |
+            cd "$ROOT_DIR"/python && pytest --cov=src --cov-report=html test
+
+
+      - name: Store artifacts
+        uses: actions/upload-artifact@v3.1.2
+        if: always()  # Always run even if the workflow is canceled manually or due to overtime.
+        # Note that `$TEST_DIR` does not work on Windows, where its equivalent is `$env:TEST_DIR`.
+        # In the following, we enquire `$TEST_DIR` by using the `env` context, which is platform independent.
+        with:
+          path: ${{ env.TEST_DIR }}/prima/python/htmlcov/*
+
+      - name: Remove the test data
+        shell: bash  # Important; otherwise, `rm -rf` will not work on Windows.
+        run: rm -rf ${{ env.TEST_DIR }}

python/README.md

@@ -0,0 +1,74 @@
+To develop, `cd` into the `src` directory and run
+
+```pip install --editable .```
+
+This will install prima locally in an editable fashion. From there you can run the examples/cobyla/cobyla_example.py (from any directory) and go from there.
+
+## Style notes
+
+- In cases where there are Python or numpy functions that can achieve the same aim (i.e. sum vs np.sum) we prefer the numpy functions.
+  - Rationale:
+    - In some cases we must use the numpy function because it has functionality the Python function doesn't, i.e. np.sum has an axis argument, whereas sum does not.
+    - In light of this, for the sake of consistency we prefer to use the numpy functions everywhere.
+- Most of the comments are copied from Fortran verbatim, except in cases where they need to modified due to specifics of the Python language. In these cases a note will be made of the difference between Fortran and Python
+  - Rationale:
+      - The main purpose of this is to keep the Python and Fortran codebases as similar as possible.
+- For determining the dimensions of an array, we exclusively use `np.size` instead of `np.shape` or `some_array.shape` or `len`
+  - Rationale:
+    - Fortran uses `SIZE` so this helps us to be as consistent with the Fortran code as possible.
+
+## A note on Fortran's `max` and `maxval` and their Python equivalents
+
+| Fortran | Python |
+|---------|--------|
+| `max`   | `np.maximum` |
+| `maxval` | `np.max` or `max` |
+
+Fortran's `max` and numpy's `maximum` accept two arguments, either of which can be a scalar or an array,
+and returns an elementwise maximum of the two arguments. In the case of a scalar and an array argument it
+returns an elementwise maximum of the scalar and each element of the array.
+
+Fortran's `maxval` takes a single argument and returns the largest item in the list, similar to numpy's `max`.
+Python's built-in `max` can either take a single argument and return the largest element, or it can take a
+variable number of arguments and return the largest of all of them.
+
+Per the style notes, prefer `np.max` over the built-in `max`.
+
+For the larger of two numbers Fortran uses `max`, and we prefer `np.maximum` over the Python `max` for consistenchy in the translation.
+
+This note applies to `min` and `minval` as well.
+
+
+## A note on indices
+
+Consider the following Fortran code
+
+```
+do i=0:5
+  print *, *
+end do
+```
+
+It can be easily and automatically translated to Python as
+
+```
+for i in range(0, 6):
+  print(i)
+```
+
+Now consider the following similar loop
+
+```
+do i=1:5
+  print *, some_array(i)
+end do
+```
+
+This can be translated to Python as
+
+```
+for i in range(1, 6):
+  print(some_array[i-1])
+```
+
+This leads to awkward Python code, since the more pythonic code would range from 0 to 5, and the indexing would be `some_array[i]`. In order to make the Python code more usable, we will attempt to write more "pythonic" code, even though that makes the translation a little bit more difficult.

python/pyproject.toml

@@ -0,0 +1,10 @@
+[project]
+name = "prima"
+version = "0.0.1"
+dependencies = [
+	"numpy"
+]
+
+[tool.setuptools.packages.find]
+where = ["src"]  # ["."] by default
+exclude = ["prima.examples*"]  # empty by default