fix typos, bump min pymatviz==0.10.1, bump ruff and fix errors

Author: janosh

.pre-commit-config.yaml

@@ -8,7 +8,7 @@ default_install_hook_types: [pre-commit, commit-msg]
 
 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.5.6
+    rev: v0.6.1
     hooks:
       - id: ruff
         args: [--fix]
@@ -57,7 +57,7 @@ repos:
         exclude: ^(site/src/figs/.+\.svelte|data/wbm/20.+\..+|site/src/(routes|figs).+\.(yaml|json)|changelog.md)$
 
   - repo: https://github.com/pre-commit/mirrors-eslint
-    rev: v9.8.0
+    rev: v9.9.0
     hooks:
       - id: eslint
         types: [file]
@@ -79,7 +79,7 @@ repos:
       - id: check-github-actions
 
   - repo: https://github.com/RobertCraigie/pyright-python
-    rev: v1.1.375
+    rev: v1.1.376
     hooks:
       - id: pyright
         args: [--level, error]

matbench_discovery/data.py

@@ -214,7 +214,7 @@ def __new__(
 
         obj._rel_path = file_path  # type: ignore[attr-defined] # noqa: SLF001
         obj._url = url  # type: ignore[attr-defined] # noqa: SLF001
-        obj._label = label  #  type: ignore[attr-defined] # noqa: SLF001
+        obj._label = label  # type: ignore[attr-defined] # noqa: SLF001
 
         return obj
 

matbench_discovery/enums.py

@@ -80,7 +80,7 @@ class Task(LabelEnum):
     S2RE = "S2RE", "structure to relaxed energy"
     S2EF = "S2EF", "structure to energy, force"
     S2EFS = "S2EFS", "structure to energy, force, stress"
-    S2EFSM = "S2EFSM" "structure to energy, force, stress, magmoms"
+    S2EFSM = "S2EFSM", "structure to energy, force, stress, magmoms"
     IP2E = "IP2E", "initial prototype to energy"
     IS2E = "IS2E", "initial structure to energy"
     # IS2RE is for models that learned a discrete version of PES like CGCNN+P

matbench_discovery/plots.py

@@ -190,10 +190,10 @@ def hist_classified_stable_vs_hull_dist(
         )[which_energy]
 
         if stability_threshold is not None:
-            for ax in [fig] if isinstance(fig, plt.Axes) else fig.flat:
-                ax.set(xlabel=xlabel, ylabel=y_label, xlim=x_lim)
+            for ax_i in [fig] if isinstance(fig, plt.Axes) else fig.flat:
+                ax_i.set(xlabel=xlabel, ylabel=y_label, xlim=x_lim)
                 label = "Stability Threshold"
-                ax.axvline(
+                ax_i.axvline(
                     stability_threshold, color="black", linestyle="--", label=label
                 )
 
@@ -228,8 +228,8 @@ def hist_classified_stable_vs_hull_dist(
         )
 
         if backend == MATPLOTLIB:
-            for ax in fig.flat if isinstance(fig, np.ndarray) else [fig]:
-                ax_acc = ax.twinx()
+            for ax_i in fig.flat if isinstance(fig, np.ndarray) else [fig]:
+                ax_acc = ax_i.twinx()
                 ax_acc.set_ylabel("Rolling Accuracy", color="darkblue")
                 ax_acc.tick_params(labelcolor="darkblue")
                 ax_acc.set(ylim=(0, 1.1))
@@ -681,14 +681,14 @@ def cumulative_metrics(
             rmse_interp = cubic_interpolate(model_range, rmse_cum[:n_pred_stable])
             dfs["RMSE"][model_name] = dict(zip(xs_model, rmse_interp(xs_model)))
 
-    for key in dfs:
+    for key, df_i in dfs.items():
+        # will be used as facet_col in plotly to split different metrics into subplots
+        df_i["metric"] = key
         # drop all-NaN rows so plotly plot x-axis only extends to largest number of
         # predicted materials by any model
-        dfs[key] = dfs[key].dropna(how="all")
-        # will be used as facet_col in plotly to split different metrics into subplots
-        dfs[key]["metric"] = key
+        dfs[key] = df_i.dropna(how="all")
 
-    df_cum = pd.concat(dfs.values())
+    df_cumu_metrics = pd.concat(dfs.values())
     # subselect rows for speed, plot has sufficient precision with 1k rows
     n_stable = sum(e_above_hull_true <= STABILITY_THRESHOLD)
 
@@ -752,7 +752,7 @@ def cumulative_metrics(
     elif backend == PLOTLY:
         n_cols = kwargs.pop("facet_col_wrap", 2)
         kwargs.setdefault("facet_col_spacing", 0.03)
-        fig = df_cum.plot(
+        fig = df_cumu_metrics.plot(
             backend=backend,
             facet_col="metric",
             facet_col_wrap=n_cols,
@@ -802,7 +802,7 @@ def cumulative_metrics(
     else:
         raise ValueError(f"Unknown {backend=}")
 
-    return fig, df_cum
+    return fig, df_cumu_metrics
 
 
 def wandb_scatter(table: wandb.Table, fields: dict[str, str], **kwargs: Any) -> None:

matbench_discovery/preds.py

@@ -72,14 +72,14 @@ class Model(Files, base_dir=f"{ROOT}/models"):
         "Wrenformer",
     )
 
-    ## Proprietary Models
+    # --- Proprietary Models
     # GNoMe
     gnome = "gnome/2023-11-01-gnome-preds-50076332.csv.gz", None, "GNoME"
 
     # MatterSim
     mattersim = "mattersim/mattersim-wbm-IS2RE.csv.gz", None, "MatterSim"
 
-    ## Miscellaneous
+    # --- Model Combos
     # # CHGNet-relaxed structures fed into MEGNet for formation energy prediction
     # chgnet_megnet = "chgnet/2023-03-06-chgnet-0.2.0-wbm-IS2RE.csv.gz", None, "CHGNet→MEGNet"
     # # M3GNet-relaxed structures fed into MEGNet for formation energy prediction

models/mattersim/test_mattersim.py

@@ -35,7 +35,7 @@
 
 
 def dummy_mattersim_calculator(
-    backbone: Literal["m3gnet", "graphomer"] = "m3gnet",
+    backbone: Literal["m3gnet", "graphormer"] = "m3gnet",
 ) -> SinglePointCalculator:
     """
     This is a dummy function that makes a MatterSim calculator
@@ -69,7 +69,7 @@ def relax_atoms_list(
     atoms_list: list[ase.Atoms],
     fmax: float = 0.01,
     steps: int = 500,
-    backbone: Literal["m3gnet", "graphomer"] = "m3gnet",
+    backbone: Literal["m3gnet", "graphormer"] = "m3gnet",
 ) -> list[ase.Atoms]:
     """
     This function relax the atoms.
@@ -88,11 +88,11 @@ def relax_atoms_list(
         else:
             atoms.info["converged"] = True
 
-        if backbone == "graphomer":
+        if backbone == "graphormer":
             # Please note that we only re-calculate the
-            # energy in the case of MatterSim(graphomer).
+            # energy in the case of MatterSim(graphormer).
             # The structure relaxation is always done with MatterSim(m3gnet).
-            calc = dummy_mattersim_calculator(backbone="graphomer")
+            calc = dummy_mattersim_calculator(backbone="graphormer")
             atoms.set_calculator(calc)
 
         relaxed_atoms_list.append(atoms)
@@ -106,7 +106,6 @@ def parse_relaxed_atoms_list_as_df(
 ) -> pd.DataFrame:
     e_form_col = "e_form_per_atom_mattersim"
 
-    ## Read pre-computed CSEs by WBM
     wbm_cse_paths = DataFiles.wbm_computed_structure_entries.path
     df_cse = pd.read_json(wbm_cse_paths).set_index(Key.mat_id)
 
@@ -167,7 +166,9 @@ def parse_single_atoms(atoms: ase.Atoms) -> tuple[str, bool, float, float, float
 
 if __name__ == "__main__":
     init_wbm_atoms_list = convert_wbm_to_atoms_list()
-    relaxed_wbm_atoms_list = relax_atoms_list(init_wbm_atoms_list, backbone="graphomer")
+    relaxed_wbm_atoms_list = relax_atoms_list(
+        init_wbm_atoms_list, backbone="graphormer"
+    )
     parse_relaxed_atoms_list_as_df(relaxed_wbm_atoms_list).to_csv(
         "mattersim-wbm-IS2RE.csv.gz"
     )

pyproject.toml

@@ -31,13 +31,10 @@ dependencies = [
   "pandas>=2.0.0",
   "plotly",
   "pymatgen",
-  "pymatviz[export-figs,df-pdf-export,df-svg-export]>=0.10.0",
+  "pymatviz[export-figs,df-pdf-export,df-svg-export]>=0.10.1",
   "scikit-learn",
   "scipy",
   "seaborn",
-  # TODO remove svgutils after next pymatviz release 0.10.1
-  "svgutils",
-  "svgwrite",
   "tqdm",
   "wandb",
 

tests/conftest.py

@@ -6,7 +6,7 @@
 from pymatviz.enums import Key
 
 
-@pytest.fixture()
+@pytest.fixture
 def dummy_struct() -> Structure:
     return Structure(
         lattice=Lattice.cubic(4.2),
@@ -15,14 +15,14 @@ def dummy_struct() -> Structure:
     )
 
 
-@pytest.fixture()
+@pytest.fixture
 def df_float() -> pd.DataFrame:
     rng = np.random.default_rng(0)
 
     return pd.DataFrame(rng.normal(size=(10, 5)), columns=[*"ABCDE"])
 
 
-@pytest.fixture()
+@pytest.fixture
 def df_mixed() -> pd.DataFrame:
     rng = np.random.default_rng(0)
 
@@ -32,7 +32,7 @@ def df_mixed() -> pd.DataFrame:
     return pd.DataFrame(dict(floats=floats, bools=bools, strings=strings))
 
 
-@pytest.fixture()
+@pytest.fixture
 def df_with_pmg_objects(dummy_struct: Structure) -> pd.DataFrame:
     # create a dummy df with a structure column on which to test (de-)serialization
     df_dummy = pd.DataFrame(dict(material_id=range(5), structure=[dummy_struct] * 5))

tests/test_data.py

@@ -47,7 +47,8 @@
 
 def test_as_dict_handler() -> None:
     class C:
-        def as_dict(self) -> dict[str, Any]:
+        @staticmethod
+        def as_dict() -> dict[str, Any]:
             return {"foo": "bar"}
 
     assert as_dict_handler(C()) == {"foo": "bar"}

tests/test_models.py

@@ -27,17 +27,17 @@ def test_model_dirs_have_metadata() -> None:
 
     for model_name, metadata in MODEL_METADATA.items():
         model_dir = metadata["model_dir"]
-        for key in required:
+        for key, val in required.items():
             assert key in metadata, f"Required {key=} missing in {model_dir}"
-            if isinstance(required[key], dict):
-                missing_keys = {*required[key]} - {*metadata[key]}  # type: ignore[misc]
+            if isinstance(val, dict):
+                missing_keys = {*val} - {*metadata[key]}
                 assert (
                     not missing_keys
                 ), f"Missing sub-keys {missing_keys} of {key=} in {model_dir}"
                 continue
 
-            err_msg = f"Invalid {key=}, expected {required[key]} in {model_dir}"
-            assert isinstance(metadata[key], required[key]), err_msg  # type: ignore[arg-type]
+            err_msg = f"Invalid {key=}, expected {val} in {model_dir}"
+            assert isinstance(metadata[key], val), err_msg  # type: ignore[arg-type]
 
         authors, date_added, mbd_version, yml_model_name, model_version, repo = (
             metadata[key] for key in list(required)[:-1]

tests/test_plots.py

@@ -38,7 +38,7 @@ def test_cumulative_metrics(
     backend: Backend,
     metrics: tuple[str, ...],
 ) -> None:
-    fig, df_metrics = cumulative_metrics(
+    fig, df_cumu_metrics = cumulative_metrics(
         e_above_hull_true=df_wbm[MbdKey.each_true],
         df_preds=df_wbm[models],
         backend=backend,
@@ -47,8 +47,8 @@ def test_cumulative_metrics(
         metrics=metrics,
     )
 
-    assert isinstance(df_metrics, pd.DataFrame)
-    assert list(df_metrics) == [*models, "metric"]
+    assert isinstance(df_cumu_metrics, pd.DataFrame)
+    assert list(df_cumu_metrics) == [*models, "metric"]
 
     if backend == MATPLOTLIB:
         assert isinstance(fig, plt.Figure)

tests/test_slurm.py

@@ -6,7 +6,6 @@
 from matbench_discovery.slurm import _get_calling_file_path, slurm_submit
 
 
-@patch.dict(os.environ, {"SLURM_JOB_ID": "1234"}, clear=True)
 @pytest.mark.parametrize("py_file_path", [None, "path/to/file.py"])
 @pytest.mark.parametrize("partition", [None, "fake-partition"])
 @pytest.mark.parametrize("time", [None, "0:0:1"])
@@ -34,7 +33,14 @@ def test_slurm_submit(
         pre_cmd=pre_cmd,
     )
 
-    slurm_vars = slurm_submit(**kwargs)  # type: ignore[arg-type]
+    slurm_submit(**kwargs)  # type: ignore[arg-type]
+
+    stdout, stderr = capsys.readouterr()
+    # check slurm_submit() did nothing in normal mode
+    assert stdout == stderr == ""
+
+    with patch.dict(os.environ, {"SLURM_JOB_ID": "1234"}, clear=True):
+        slurm_vars = slurm_submit(**kwargs)  # type: ignore[arg-type]
     expected_slurm_vars = dict(slurm_job_id="1234", slurm_flags="--foo")
     if time is not None:
         expected_slurm_vars["slurm_timelimit"] = time
@@ -44,15 +50,12 @@ def test_slurm_submit(
         expected_slurm_vars["pre_cmd"] = pre_cmd
     assert slurm_vars == expected_slurm_vars
 
-    stdout, stderr = capsys.readouterr()
-    # check slurm_submit() did nothing in normal mode
-    assert stderr == stderr == ""
-
     # check slurm_submit() prints cmd and calls subprocess.run() in submit mode
     with (
         pytest.raises(SystemExit),
         patch("sys.argv", ["slurm-submit"]),
         patch("matbench_discovery.slurm.subprocess.run") as mock_subprocess_run,
+        patch.dict(os.environ, {"SLURM_JOB_ID": "1234"}, clear=True),
     ):
         slurm_submit(**kwargs)  # type: ignore[arg-type]