Merge pull request #63 from matteobecchi/mega_upgrade

Removing all version limitations. Dynsight should now work on Python >= 3.10. Thank you @matteobecchi

.github/workflows/tests.yaml

@@ -8,16 +8,24 @@ on:
 jobs:
   ruff:
     runs-on: ubuntu-22.04
+    strategy:
+      matrix:
+        python-version: ['3.10', '3.13' ]
+    name: ruff on ${{ matrix.python-version }}
     steps:
       - uses: actions/checkout@v3
       - uses: actions/setup-python@v4
         with:
-          python-version: "3.10"
+          python-version: ${{ matrix.python-version }}
           cache: "pip"
       - run: pip install -e '.[dev]'
       - run: ruff check .
   mypy:
     runs-on: ubuntu-22.04
+    strategy:
+      matrix:
+        python-version: ['3.10', '3.13' ]
+    name: mypy on ${{ matrix.python-version }}
     steps:
       - uses: actions/checkout@v3
       - uses: actions/setup-python@v4
@@ -28,6 +36,10 @@ jobs:
       - run: mypy .
   ruff-format:
     runs-on: ubuntu-22.04
+    strategy:
+      matrix:
+        python-version: ['3.10', '3.13' ]
+    name: ruff-format on ${{ matrix.python-version }}
     steps:
       - uses: actions/checkout@v3
       - uses: actions/setup-python@v4
@@ -38,6 +50,10 @@ jobs:
       - run: ruff format --check .
   pytest:
     runs-on: ubuntu-22.04
+    strategy:
+      matrix:
+        python-version: ['3.10', '3.13' ]
+    name: pytest on ${{ matrix.python-version }}
     steps:
       - uses: actions/checkout@v3
       - uses: actions/setup-python@v4
@@ -48,6 +64,10 @@ jobs:
       - run: pytest --cov=src --cov-report term-missing
   doctest:
     runs-on: ubuntu-22.04
+    strategy:
+      matrix:
+        python-version: ['3.10', '3.13' ]
+    name: doctest on ${{ matrix.python-version }}
     steps:
       - uses: actions/checkout@v3
       - uses: actions/setup-python@v4

pyproject.toml

@@ -11,13 +11,13 @@ maintainers = [
 ]
 
 dependencies = [
-    "numpy < 1.24.0",          # Needed right now, to remove in the future
-    "dscribe >1.2.0, <=1.2.2",
+    "numpy",
+    "dscribe",
     "tropea-clustering",
     "MDAnalysis",
 ]
 # Set by cpctools.
-requires-python = "<3.11,>=3.8"
+requires-python = ">=3.8"
 dynamic = ["version"]
 readme = "README.rst"
 description = "Simplifies analysis of Molecular Dynamics simulations."

src/dynsight/_internal/analysis/entropy.py

@@ -52,7 +52,7 @@ def compute_data_entropy(
         .. testcode:: shannon1-test
             :hide:
 
-            assert data_entropy == 0.9993954419344714
+            assert np.isclose(data_entropy, 0.9993954419344714)
     """
     counts, _ = np.histogram(
         data,
@@ -108,7 +108,7 @@ def compute_entropy_gain(
         .. testcode:: shannon2-test
             :hide:
 
-            assert entropy_gain == 0.0010065005804883983
+            assert np.isclose(entropy_gain, 0.0010065005804883983)
     """
     if data.shape[0] != labels.shape[0]:
         msg = (

src/dynsight/_internal/analysis/spatial_average.py

@@ -4,18 +4,19 @@
 
 if TYPE_CHECKING:
     import MDAnalysis
+    from numpy.typing import NDArray
 
 import ctypes
 from multiprocessing import Array, Pool
 
 import numpy as np
 from MDAnalysis.analysis.distances import distance_array
 
-array: np.ndarray[float, Any]
+array: NDArray[np.float64]
 
 
 def initworker(
-    shared_array: np.ndarray[float, Any],
+    shared_array: NDArray[np.float64],
     shape: int,
     dtype: tuple[float, Any],
 ) -> None:
@@ -27,7 +28,7 @@ def initworker(
 
 def processframe(
     args: tuple[MDAnalysis.Universe, MDAnalysis.AtomGroup, float, int, bool],
-) -> tuple[int, np.ndarray[float, Any]]:
+) -> tuple[int, NDArray[np.float64]]:
     universe, selection, cutoff, frame, is_vector = args
     universe.trajectory[frame]
     distances = distance_array(
@@ -59,12 +60,12 @@ def processframe(
 
 def spatialaverage(
     universe: MDAnalysis.Universe,
-    descriptor_array: np.ndarray[float, Any],
+    descriptor_array: NDArray[np.float64],
     selection: str,
     cutoff: float,
     traj_cut: int = 0,
     num_processes: int = 4,
-) -> np.ndarray[float, Any]:
+) -> NDArray[np.float64]:
     """Compute spatially averaged descriptor values over neighboring particles.
 
     This function takes a molecular dynamics (MD) simulation stored in an

src/dynsight/_internal/analysis/time_correlations.py

@@ -2,15 +2,18 @@
 
 from __future__ import annotations
 
-from typing import Any
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from numpy.typing import NDArray
 
 import numpy as np
 
 
 def self_time_correlation(
-    data: np.ndarray[float, Any],
+    data: NDArray[np.float64],
     max_delay: int | None = None,
-) -> tuple[np.ndarray[float, Any], np.ndarray[float, Any]]:
+) -> tuple[NDArray[np.float64], NDArray[np.float64]]:
     """Computes the mean self time correlation function for time-series.
 
     Takes as input an array of shape (n_particles, n_frames), where the
@@ -92,9 +95,9 @@ def self_time_correlation(
 
 
 def cross_time_correlation(
-    data: np.ndarray[float, Any],
+    data: NDArray[np.float64],
     max_delay: int | None = None,
-) -> tuple[np.ndarray[float, Any], np.ndarray[float, Any]]:
+) -> tuple[NDArray[np.float64], NDArray[np.float64]]:
     """Computes the mean cross time correlation function for time-series.
 
     Takes as input an array of shape (n_particles, n_frames), where the

src/dynsight/_internal/lens/lens.py

@@ -1,9 +1,10 @@
 from __future__ import annotations
 
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING
 
 if TYPE_CHECKING:
     from MDAnalysis import AtomGroup, Universe
+    from numpy.typing import NDArray
 
 import numpy as np
 from MDAnalysis.lib.NeighborSearch import AtomNeighborSearch
@@ -79,10 +80,10 @@ def list_neighbours_along_trajectory(
 def neighbour_change_in_time(
     neigh_list_per_frame: list[list[AtomGroup]],
 ) -> tuple[
-    np.ndarray[float, Any],
-    np.ndarray[int, Any],
-    np.ndarray[float, Any],
-    np.ndarray[float, Any],
+    NDArray[np.float64],
+    NDArray[np.int64],
+    NDArray[np.float64],
+    NDArray[np.float64],
 ]:
     """Return, listed per atom, the LENS values at each frame.
 
@@ -143,7 +144,7 @@ def neighbour_change_in_time(
     # this is the number of common NN between frames
     lensarray = np.zeros((nat, nframes))
     # this is the number of NN at that frame
-    numberofneighs = np.zeros((nat, nframes))
+    numberofneighs = np.zeros((nat, nframes), dtype=int)
     # this is the numerator of LENS
     lensnumerators = np.zeros((nat, nframes))
     # this is the denominator of lens

src/dynsight/_internal/soapify/saponify.py

@@ -6,6 +6,7 @@
 
 if TYPE_CHECKING:
     import MDAnalysis
+    from numpy.typing import NDArray
 
 import numpy as np
 from ase import Atoms
@@ -21,7 +22,7 @@ def saponify_trajectory(
     soap_respectpbc: bool = True,
     n_core: int = 1,
     **soapkwargs: Any,
-) -> np.ndarray[float, Any]:
+) -> NDArray[np.float64]:
     """Calculate the SOAP fingerprints for each atom in a MDA universe.
 
     * Author: Simone Martino
@@ -108,10 +109,10 @@ def saponify_trajectory(
 
 
 def fill_soap_vector_from_dscribe(
-    soapfromdscribe: np.ndarray[float, Any],
+    soapfromdscribe: NDArray[np.float64],
     lmax: int = 8,
     nmax: int = 8,
-) -> np.ndarray[float, Any]:
+) -> NDArray[np.float64]:
     """Return the SOAP power spectrum from dscribe results.
 
     * Author: Matteo Becchi

src/dynsight/_internal/timesoap/timesoap.py

@@ -2,14 +2,17 @@
 
 from __future__ import annotations
 
-from typing import Any
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from numpy.typing import NDArray
 
 import numpy as np
 
 
 def normalize_soap(
-    soaptrajectory: np.ndarray[float, Any],
-) -> np.ndarray[float, Any]:
+    soaptrajectory: NDArray[np.float64],
+) -> NDArray[np.float64]:
     """Returns the SOAP spectra normalized to unitary length.
 
     Parameters:
@@ -58,9 +61,9 @@ def normalize_soap(
 
 
 def soap_distance(
-    v_1: np.ndarray[float, Any],
-    v_2: np.ndarray[float, Any],
-) -> np.ndarray[float, Any]:
+    v_1: NDArray[np.float64],
+    v_2: NDArray[np.float64],
+) -> NDArray[np.float64]:
     r"""Computes the Kernel SOAP distance between two SOAP spectra.
 
     The SOAP distance is calculated with:
@@ -86,7 +89,7 @@ def soap_distance(
             SOAP spectra.
 
     Returns:
-        np.ndarray[float, Any] : the SOAP distances between the input spectra.
+        NDArray[np.float64] : the SOAP distances between the input spectra.
 
     Example:
 
@@ -127,9 +130,9 @@ def soap_distance(
 
 
 def timesoap(
-    soaptrajectory: np.ndarray[float, Any],
+    soaptrajectory: NDArray[np.float64],
     delay: int = 1,
-) -> np.ndarray[float, Any]:
+) -> NDArray[np.float64]:
     """Performs the 'timeSOAP' analysis on the given SOAP trajectory.
 
     timeSOAP was developed by Cristina Caurso. See for reference the paper
@@ -144,7 +147,7 @@ def timesoap(
             Default is 1.
 
     Returns:
-        np.ndarray[float, Any]
+        NDArray[np.float64]
             Values of timesoap of each particle at each frame.
             Has shape (n_particles, n_frames - 1).
 

tests/analysis/test_shannon.py

@@ -36,9 +36,13 @@ def test_output_files(original_wd: Path) -> None:  # noqa: ARG001
         os.chdir(temp_dir)
 
         # Test the use of the function computing entropy
+
+        # This is necessary because of type checking:
+        data_min = float(np.min(random_data))
+        data_max = float(np.max(random_data))
         data_entropy = dynsight.analysis.compute_data_entropy(
             random_data,
-            data_range=(np.min(random_data), np.max(random_data)),
+            data_range=(data_min, data_max),
             n_bins=20,
         )