Merge pull request #13 from keurfonluu/devel

Add option verbosity and modernize repository

.github/workflows/ci.yml

@@ -8,7 +8,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.6, 3.7, 3.8, 3.9]
+        python-version: ["3.7", "3.8", "3.9", "3.10", "3.11"]
     steps:
     - uses: actions/setup-python@v2
       with:

.github/workflows/doc.yml

@@ -0,0 +1,48 @@
+name: Build documentation
+
+on:
+  push:
+    branches:
+      - master
+  pull_request:
+
+jobs:
+  docs:
+
+    runs-on: ubuntu-latest
+    env:
+      PYVISTA_OFF_SCREEN: 'True'
+      ALLOW_PLOTTING: true
+      SHELLOPTS: 'errexit:pipefail'
+
+    steps:
+      - uses: nschloe/action-cached-lfs-checkout@v1
+      - uses: actions/setup-python@v4
+        with:
+            python-version: '3.9'
+      - name: Install OSMesa VTK
+        run: |
+          sudo apt update
+          sudo apt install libosmesa6-dev libgl1-mesa-dev retry --option="APT::Acquire::Retries=3"
+          pip install https://github.com/pyvista/pyvista-wheels/raw/main/vtk-osmesa-9.1.0-cp39-cp39-linux_x86_64.whl
+      - name: Install libGLU for pygmsh
+        run: |
+          sudo apt install libglu1
+      - name: Install toughio
+        run: |
+          git lfs pull
+          pip install -e .[full]
+      - name: Install dependencies
+        run: |
+          pip install -r doc/requirements.txt
+      - name: Sphinx build
+        run: |
+          sphinx-build -b html doc/source doc/build
+      - name: Deploy
+        uses: peaceiris/actions-gh-pages@v3
+        if: ${{ github.event_name == 'push' }}
+        with:
+          publish_branch: gh-pages
+          github_token: ${{ secrets.GITHUB_TOKEN }}
+          publish_dir: doc/build/
+          force_orphan: true

README.rst

@@ -1,7 +1,7 @@
 stochopy
 ========
 
-|License| |Stars| |Pyversions| |Version| |Downloads| |Code style: black| |Codacy Badge| |Codecov| |Build| |Travis| |DOI|
+|License| |Stars| |Pyversions| |Version| |Downloads| |Code style: black| |Codacy Badge| |Codecov| |Build| |Docs| |DOI|
 
 **stochopy** provides functions for sampling or optimizing objective functions with or without constraints. Its API is directly inspired by **scipy**'s own optimization submodule which should make the switch from one module to another straightforward.
 
@@ -142,8 +142,8 @@ Related projects
 .. |DOI| image:: https://zenodo.org/badge/DOI/10.5281/zenodo.4058008.svg?style=flat
    :target: https://doi.org/10.5281/zenodo.4058008
 
-.. |Build| image:: https://img.shields.io/github/workflow/status/keurfonluu/stochopy/Python%20package
+.. |Build| image:: https://img.shields.io/github/workflow/status/keurfonluu/stochopy/ci.yml
    :target: https://github.com/keurfonluu/stochopy
 
-.. |Travis| image:: https://img.shields.io/travis/com/keurfonluu/stochopy/master?label=docs
+.. |Docs| image:: https://img.shields.io/github/actions/workflow/status/keurfonluu/stochopy/doc.yml?label=docs
    :target: https://keurfonluu.github.io/stochopy/

doc/requirements.txt

@@ -4,4 +4,4 @@ sphinx-argparse
 sphinx-gallery
 pydata-sphinx-theme
 sphinxcontrib-bibtex
--e git://github.com/Naeka/pybtex-apa-style.git#egg=pybtex-apa-style
+pybtex-apa-style

requirements-dev.txt

@@ -1,6 +1,8 @@
 black
+build
 docformatter
 invoke
 isort
 pytest
-pytest-cov
+pytest-cov
+twine

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = stochopy
-version = 2.2.0
+version = file: stochopy/VERSION
 author = Keurfon Luu
 email = keurfonluu@outlook.com
 description = Python library for stochastic numerical optimization
@@ -16,18 +16,23 @@ classifiers =
     License :: OSI Approved :: BSD License
     Operating System :: OS Independent
     Development Status :: 5 - Production/Stable
-    Programming Language :: Python :: 3.6
     Programming Language :: Python :: 3.7
     Programming Language :: Python :: 3.8
     Programming Language :: Python :: 3.9
+    Programming Language :: Python :: 3.10
+    Programming Language :: Python :: 3.11
 
 [options]
 packages = find:
 install_requires =
     importlib_metadata
     numpy
     joblib >= 0.12
-python_requires = >= 3.6
+python_requires = >= 3.7
 setup_requires =
     setuptools >= 42
-    wheel
+    wheel
+
+[options.package_data]
+stochopy =
+    VERSION

stochopy/VERSION

@@ -0,0 +1 @@
+2.3.0

stochopy/__about__.py

@@ -1,9 +1,9 @@
-try:
-    from importlib import metadata
-except ImportError:
-    import importlib_metadata as metadata
+import pathlib
 
-try:
-    __version__ = metadata.version("stochopy")
-except Exception:
-    __version__ = "unknown"
+__all__ = [
+    "__version__",
+]
+
+
+with open(f"{pathlib.Path(__file__).parent}/VERSION") as f:
+    __version__ = f.readline().strip()

stochopy/optimize/_helpers.py

@@ -27,8 +27,8 @@ class OptimizeResult(BaseResult):
     nit : int
         Number of iterations performed by the optimizer.
 
-    Notes
-    -----
+    Note
+    ----
     There may be additional attributes not listed above depending of the specific solver.
 
     """
@@ -70,7 +70,8 @@ def minimize(fun, bounds, x0=None, args=(), method="de", options=None, callback=
 
          - maxiter (int): maximum number of iterations to perform
          - seed (int or None): seed for random number generator
-         - return_all (bool): set to True to return an array of all the solutions at each iteration.
+         - return_all (bool): set to True to return an array of all the solutions at each iteration
+         - verbosity: fraction of population to consider in `return_all`
 
     callback : callable or None, optional, default None
         Called after each iteration. It is a callable with the signature ``callback(X, OptimizeResult state)``, where ``X`` is the current population and ``state`` is a partial :class:`stochopy.optimize.OptimizeResult` object with the same fields as the ones from the return (except ``"success"``, ``"status"`` and ``"message"``).

stochopy/optimize/cmaes/_cmaes.py

@@ -25,6 +25,7 @@ def minimize(
     workers=1,
     backend=None,
     return_all=False,
+    verbosity=1.0,
     callback=None,
 ):
     """
@@ -70,7 +71,9 @@ def minimize(
          - 'mpi': use MPI (uses :mod:`mpi4py`)
 
     return_all : bool, optional, default False
-        Set to True to return an array with shape (``nit``, ``popsize``, ``ndim``) of all the solutions at each iteration.
+        Set to True to return an array with shape (``nit``, ``verbosity`` * ``popsize``, ``ndim``) of all the solutions at each iteration.
+    verbosity : float, optional, default 1.0
+        Fraction of population to consider in `return_all`. If 0.0, returns the best solution at each iteration.
     callback : callable or None, optional, default None
         Called after each iteration. It is a callable with the signature ``callback(X, OptimizeResult state)``, where ``X`` is the current population and ``state`` is a partial :class:`stochopy.optimize.OptimizeResult` object with the same fields as the ones from the return (except ``"success"``, ``"status"`` and ``"message"``).
 
@@ -129,6 +132,7 @@ def minimize(
         xtol,
         ftol,
         return_all,
+        verbosity,
         callback,
     )
     res = cmaes(fun, args, True, workers, backend, *optargs)
@@ -153,6 +157,7 @@ def cmaes(
     xtol,
     ftol,
     return_all,
+    verbosity,
     callback,
 ):
     """Optimize with CMA-ES."""
@@ -197,16 +202,17 @@ def cmaes(
     D = np.ones(ndim)
     C = np.eye(ndim)
     invsqrtC = np.eye(ndim)
-    chind = np.sqrt(ndim) * (1.0 - 1.0 / (4.0 * ndim) + 1.0 / (21.0 * ndim ** 2))
+    chind = np.sqrt(ndim) * (1.0 - 1.0 / (4.0 * ndim) + 1.0 / (21.0 * ndim**2))
 
     # Initialize boundaries weights
     bnd_weights = np.zeros(ndim)
     dfithist = np.ones(1)
 
     # Initialize arrays
     if return_all:
-        xall = np.empty((maxiter, popsize, ndim))
-        funall = np.empty((maxiter, popsize))
+        nout = int(np.ceil(verbosity * popsize))
+        xall = np.empty((maxiter, max(1, nout), ndim))
+        funall = np.empty((maxiter, max(1, nout)))
 
     # (mu, lambda)-CMA-ES
     nfev = 0
@@ -253,8 +259,14 @@ def cmaes(
         nfev += popsize
 
         if return_all:
-            xall[it - 1] = unstandardize(arxvalid)
-            funall[it - 1] = arfitness.copy()
+            if nout > 0:
+                xall[it - 1] = unstandardize(arxvalid[:nout])
+                funall[it - 1] = arfitness[:nout].copy()
+
+            else:
+                idx = arfitness.argmin()
+                xall[it - 1] = unstandardize(arxvalid[idx])
+                funall[it - 1] = arfitness[idx].copy()
 
         # Sort by fitness and compute weighted mean into xmean
         arindex = np.argsort(arfitness)

stochopy/optimize/cmaes/_constraints.py

@@ -33,7 +33,7 @@ def Penalize(
 
     # Get delta fitness values
     perc = np.percentile(arfitness, [25.0, 75.0])
-    delta = (perc[1] - perc[0]) / ndim / diagC.mean() / sigma ** 2
+    delta = (perc[1] - perc[0]) / ndim / diagC.mean() / sigma**2
 
     # Catch non-sensible values
     if delta == 0:

stochopy/optimize/cpso/_cpso.py

@@ -28,6 +28,7 @@ def minimize(
     workers=1,
     backend=None,
     return_all=False,
+    verbosity=1.0,
     callback=None,
 ):
     """
@@ -79,7 +80,9 @@ def minimize(
          - 'mpi': use MPI (uses :mod:`mpi4py`)
 
     return_all : bool, optional, default False
-        Set to True to return an array with shape (``nit``, ``popsize``, ``ndim``) of all the solutions at each iteration.
+        Set to True to return an array with shape (``nit``, ``verbosity`` * ``popsize``, ``ndim``) of all the solutions at each iteration.
+    verbosity : float, optional, default 1.0
+        Fraction of population to consider in `return_all`. If 0.0, returns the best solution at each iteration.
     callback : callable or None, optional, default None
         Called after each iteration. It is a callable with the signature ``callback(X, OptimizeResult state)``, where ``X`` is the current population and ``state`` is a partial :class:`stochopy.optimize.OptimizeResult` object with the same fields as the ones from the return (except ``"success"``, ``"status"`` and ``"message"``).
 
@@ -168,6 +171,7 @@ def minimize(
         xtol,
         ftol,
         return_all,
+        verbosity,
         callback,
     )
     res = cpso(fun, args, sync, workers, backend, *optargs)
@@ -194,6 +198,7 @@ def cpso(
     xtol,
     ftol,
     return_all,
+    verbosity,
     callback,
 ):
     """Optimize with CPSO."""
@@ -226,10 +231,19 @@ def cpso(
 
     # Initialize arrays
     if return_all:
-        xall = np.empty((maxiter, popsize, ndim))
-        funall = np.empty((maxiter, popsize))
-        xall[0] = X.copy()
-        funall[0] = pfit.copy()
+        nout = int(np.ceil(verbosity * popsize))
+
+        if nout > 0:
+            xall = np.empty((maxiter, nout, ndim))
+            funall = np.empty((maxiter, nout))
+            xall[0] = X[:nout].copy()
+            funall[0] = pfit[:nout].copy()
+
+        else:
+            xall = np.empty((maxiter, 1, ndim))
+            funall = np.empty((maxiter, 1))
+            xall[0] = gbest
+            funall[0] = gfit
 
     # First iteration for callback
     if callback is not None:
@@ -269,8 +283,14 @@ def cpso(
         )
 
         if return_all:
-            xall[it - 1] = X.copy()
-            funall[it - 1] = pfit.copy()
+            if nout > 0:
+                xall[it - 1] = X[:nout].copy()
+                funall[it - 1] = pfit[:nout].copy()
+
+            else:
+                idx = pfit.argmin()
+                xall[it - 1] = X[idx].copy()
+                funall[it - 1] = pfit[idx].copy()
 
         converged = status is not None