save units in attrs of H5 (#90)

* save units in attrs of H5

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* hardcode PBC and 3D data

* only create groups when required

* fix time

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* test positions

* remove `initialize_database_groups`

* bump version to 0.2.0

* test multiple python versions

* bump version test

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

.github/workflows/pytest.yaml

@@ -9,30 +9,39 @@ on:
     - cron: '14 3 * * 1'  # at 03:14 on Monday.
 
 jobs:
-  test:
-
-    runs-on: ubuntu-latest
+  pytest:
+    runs-on: ${{ matrix.os }}
     strategy:
       fail-fast: false
       matrix:
         python-version:
+          - "3.12"
+          - "3.11"
           - "3.10"
+          - "3.9"
+        os:
+          - ubuntu-latest
 
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v4
       - name: Set up Python ${{ matrix.python-version }}
-        uses: actions/setup-python@v2
+        uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
       - name: Install Poetry
-        uses: abatilo/actions-poetry@v2
+        uses: snok/install-poetry@v1
         with:
-          poetry-version: 1.3.2
+          virtualenvs-create: true
+          virtualenvs-in-project: true
+      - name: Install package
+        run: |
+          poetry install --no-interaction --without=notebook --all-extras
       - name: Install package
         run: |
           poetry install --without=notebook --all-extras
       - name: Pytest
         run: |
+          poetry run python --version
           poetry run coverage run -m pytest
           poetry run coverage lcov
       - name: Coveralls

README.md

@@ -63,7 +63,6 @@ import ase
 atoms: list[ase.Atoms]
 
 db = znh5md.io.DataWriter(filename="db.h5")
-db.initialize_database_groups()
 
 db.add(znh5md.io.AtomsReader(atoms)) # or znh5md.io.ChemfilesReader
 

examples/ase.ipynb

@@ -69,7 +69,6 @@
    "source": [
     "%%time\n",
     "db = znh5md.io.DataWriter(filename=\"db.h5\")\n",
-    "db.initialize_database_groups()\n",
     "db.add(znh5md.io.AtomsReader(atoms_list, step=1, time=0.1))"
    ]
   },
@@ -253,7 +252,6 @@
    ],
    "source": [
     "db = znh5md.io.DataWriter(filename=\"from_file.h5\")\n",
-    "db.initialize_database_groups()\n",
     "db.add(znh5md.io.ASEFileReader(\"traj.xyz\", step=1, time=0.1))"
    ]
   },

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "znh5md"
-version = "0.1.9"
+version = "0.2.0"
 description = "High Performance Interface for H5MD Trajectories"
 authors = ["zincwarecode <zincwarecode@gmail.com>"]
 license = "Apache-2.0"
@@ -22,6 +22,7 @@ pre-commit = "^2.20.0"
 coverage = "^7.1.0"
 pytest-profiling = "^1.7.0"
 pytest-benchmark = "^4.0.0"
+mdanalysis = "^2"
 
 [tool.poetry.group.notebook.dependencies]
 jupyterlab = "^4"

tests/test_ASEH5MD.py

@@ -29,7 +29,6 @@ def test_get_slice(tmp_path, atoms_list, remove_calc):
             atoms.calc = None
 
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
     db.add(znh5md.io.AtomsReader(atoms_list))
 
     traj = znh5md.ASEH5MD("db.h5")
@@ -56,7 +55,6 @@ def test_request_missing_properties(tmp_path, atoms_list, remove_calc):
             atoms.calc = None
 
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
 
     if remove_calc:
         with pytest.raises(RuntimeError):
@@ -77,7 +75,6 @@ def test_skip_property(tmp_path, atoms_list):
     os.chdir(tmp_path)
 
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
 
     atoms_list[-1].arrays.pop("momenta")
     atoms_list[-1].get_momenta()

tests/test_benchmarks.py

@@ -6,15 +6,13 @@
 def test_bench_DataWriter(tmp_path, atoms_list, benchmark):
     os.chdir(tmp_path)
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
     reader = znh5md.io.AtomsReader(atoms_list)
     benchmark(db.add, reader)
 
 
 def test_bench_ASEH5MD(tmp_path, atoms_list, benchmark):
     os.chdir(tmp_path)
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
     reader = znh5md.io.AtomsReader(atoms_list)
     db.add(reader)
 

tests/test_custom_file_handle.py

@@ -12,7 +12,6 @@ def test_AtomsReader(tmp_path, atoms_list):
     print(tmp_path)
 
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
 
     reader = znh5md.io.AtomsReader(atoms_list, frames_per_chunk=10)
     db.add(reader)

tests/test_mdanalysis.py

@@ -0,0 +1,41 @@
+"""Use the MDAnalysis library to read H5 files and check compliance with the H5MD standard."""
+import pathlib
+
+import ase.build
+import MDAnalysis as mda
+import numpy as np
+import pytest
+from MDAnalysis.coordinates.H5MD import H5MDReader
+
+import znh5md
+
+
+@pytest.fixture
+def trajectory() -> list[ase.Atoms]:
+    """Generate ase.Atoms objects that moves linearly in space."""
+    water = ase.build.molecule("H2O")
+    atoms_list = [water]
+    while len(atoms_list) < 100:
+        atoms = atoms_list[-1].copy()
+        atoms.positions += [0.1, 0.1, 0.1]
+        atoms_list.append(atoms)
+    return atoms_list
+
+
+@pytest.fixture
+def h5_trajectory(tmp_path, trajectory) -> pathlib.Path:
+    """Write the trajectory to an H5 file."""
+    filename = tmp_path / "trajectory.h5"
+    db = znh5md.io.DataWriter(filename=filename)
+    reader = znh5md.io.AtomsReader(trajectory)
+    db.add(reader)
+    return filename
+
+
+def test_read_h5md(h5_trajectory, trajectory):
+    u = mda.Universe.empty(n_atoms=3, trajectory=True)
+    reader = H5MDReader(h5_trajectory)
+    u.trajectory = reader
+    assert len(u.trajectory) == 100
+    for ref, ts in zip(trajectory, u.trajectory):
+        assert np.allclose(ref.positions, ts.positions)

tests/test_readers.py

@@ -17,7 +17,6 @@ def test_AtomsReader(tmp_path, reader, atoms_list, use_add):
     print(tmp_path)
 
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
 
     if reader == znh5md.io.AtomsReader:
         inputs = atoms_list
@@ -72,7 +71,6 @@ def test_ChemfilesReader(tmp_path, atoms_list, frames_per_chunk):
     print(tmp_path)
 
     db = znh5md.io.DataWriter(filename="db.h5")
-    db.initialize_database_groups()
 
     inputs = "traj.xyz"
     ase.io.write(inputs, atoms_list)

tests/test_writing.py

@@ -13,7 +13,6 @@ def test_DataWriter(tmp_path, atoms_list):
     os.chdir(tmp_path)
 
     db = DataWriter(filename="db.h5")
-    db.initialize_database_groups()
 
     reader = AtomsReader(atoms_list)
 

tests/test_znh5md.py

@@ -5,7 +5,7 @@
 
 
 def test_version():
-    assert znh5md.__version__ == "0.1.9"
+    assert znh5md.__version__ == "0.2.0"
 
 
 def test_shape(example_h5):

znh5md/cli.py

@@ -36,9 +36,6 @@ def convert(file: str, db_file: str):
     import znh5md
 
     db = znh5md.io.DataWriter(db_file)
-    if not pathlib.Path(db_file).exists():
-        db.initialize_database_groups()
-
     db.add(znh5md.io.ASEFileReader(file))
 
 

znh5md/format/__init__.py

@@ -67,6 +67,25 @@ def decode_boundary(value) -> np.ndarray:
 OBSERVABLES_GRP = [GRP.energy, GRP.stress]
 
 
+@dataclasses.dataclass
+class StepTimeChunkUnits:
+    value: str
+    time: str
+
+
+# We store everything in ASE units
+UNITS_PER_GRP = {
+    GRP.position: StepTimeChunkUnits(value="Angstrom", time="fs"),
+    GRP.velocity: StepTimeChunkUnits(value="Angstrom/fs", time="fs"),
+    GRP.forces: StepTimeChunkUnits(value="eV/Angstrom", time="fs"),
+    GRP.energy: StepTimeChunkUnits(value="eV", time="fs"),
+    GRP.stress: StepTimeChunkUnits(value="GPa", time="fs"),
+    GRP.momentum: StepTimeChunkUnits(value="eV/fs", time="fs"),
+    GRP.edges: StepTimeChunkUnits(value="Angstrom", time="fs"),
+    GRP.boundary: StepTimeChunkUnits(value="Angstrom", time="fs"),
+}
+
+
 @dataclasses.dataclass
 class FormatHandler:
     filename: str