152 ruff apply unsafe code fixes for code basis (#153)

Author: c-w-feldmann

README.md

@@ -122,7 +122,7 @@ physchem_matrix
 #                [88.065, 20.23 ,  1.   ]])
 ```
 
-MolPipeline provides further features and descriptors from RDKit, 
+MolPipeline provides further features and descriptors from RDKit,
 for example Morgan (binary/count) fingerprints and MACCS keys.
 See the [04_feature_calculation notebook](notebooks/04_feature_calculation.ipynb) for more examples.
 

molpipeline/estimators/chemprop/component_wrapper.py

@@ -31,9 +31,7 @@
 from chemprop.nn.predictors import MveFFN as _MveFFN
 from chemprop.nn.predictors import RegressionFFN as _RegressionFFN
 from chemprop.nn.predictors import SpectralFFN as _SpectralFFN
-from chemprop.nn.predictors import (
-    _FFNPredictorBase as _Predictor,  # pylint: disable=protected-access
-)
+from chemprop.nn.predictors import _FFNPredictorBase as _Predictor
 from chemprop.nn.transforms import UnscaleTransform
 from chemprop.nn.utils import Activation, get_activation_function
 from sklearn.base import BaseEstimator
@@ -477,7 +475,7 @@ def reinitialize_network(self) -> Self:
             # pylint: disable=protected-access
             self.metrics = [self.predictor._T_default_metric, self.criterion]
         else:
-            self.metrics = list(self.metric_list) + [self.criterion]
+            self.metrics = [*list(self.metric_list), self.criterion]
 
         return self
 

molpipeline/estimators/chemprop/featurizer_wrapper/graph_wrapper.py

@@ -20,7 +20,8 @@ class SimpleMoleculeMolGraphFeaturizer(_SimpleMoleculeMolGraphFeaturizer):
     extra_bond_fdim: InitVar[int]
 
     def get_params(
-        self, deep: bool = True  # pylint: disable=unused-argument
+        self,
+        deep: bool = True,  # pylint: disable=unused-argument
     ) -> dict[str, InitVar[int]]:
         """Get parameters for the featurizer.
 

molpipeline/estimators/chemprop/models.py

@@ -149,7 +149,8 @@ def __sklearn_tags__(self) -> Tags:
         return tags
 
     def _predict(
-        self, X: MoleculeDataset  # pylint: disable=invalid-name
+        self,
+        X: MoleculeDataset,  # pylint: disable=invalid-name
     ) -> npt.NDArray[np.float64]:
         """Predict the labels.
 
@@ -207,7 +208,8 @@ def fit(
         return super().fit(X, y)
 
     def predict(
-        self, X: MoleculeDataset  # pylint: disable=invalid-name
+        self,
+        X: MoleculeDataset,  # pylint: disable=invalid-name
     ) -> npt.NDArray[np.float64]:
         """Predict the output.
 
@@ -234,7 +236,8 @@ def predict(
 
     @available_if(_is_classifier)
     def predict_proba(
-        self, X: MoleculeDataset  # pylint: disable=invalid-name
+        self,
+        X: MoleculeDataset,  # pylint: disable=invalid-name
     ) -> npt.NDArray[np.float64]:
         """Predict the probabilities.
 

molpipeline/estimators/chemprop/neural_fingerprint.py

@@ -85,7 +85,8 @@ def fit(
         return super().fit(X, y)
 
     def transform(
-        self, X: MoleculeDataset  # pylint: disable=invalid-name
+        self,
+        X: MoleculeDataset,  # pylint: disable=invalid-name
     ) -> npt.NDArray[np.float64]:
         """Transform the input.
 

molpipeline/estimators/similarity_transformation.py

@@ -5,7 +5,7 @@
 from typing import Any
 
 try:
-    from typing import Self  # pylint: disable=no-name-in-module
+    from typing import Self
 except ImportError:
     from typing_extensions import Self
 
@@ -92,7 +92,8 @@ def fit(
         return self
 
     def transform(
-        self, X: npt.NDArray[np.float64] | csr_matrix  # pylint: disable=invalid-name
+        self,
+        X: npt.NDArray[np.float64] | csr_matrix,  # pylint: disable=invalid-name
     ) -> npt.NDArray[np.float64]:
         """Transform the data.
 
@@ -112,7 +113,7 @@ def transform(
 
     def fit_transform(
         self,
-        X: npt.NDArray[np.float64] | csr_matrix,  # pylint: disable=invalid-name
+        X: npt.NDArray[np.float64] | csr_matrix,
         y: npt.NDArray[np.float64] | None = None,
         **fit_params: Any,
     ) -> npt.NDArray[np.float64]:

molpipeline/experimental/explainability/explainer.py

@@ -162,7 +162,9 @@ class AbstractSHAPExplainer(abc.ABC):  # pylint: disable=too-few-public-methods
 
     @abc.abstractmethod
     def explain(
-        self, X: Any, **kwargs: Any  # pylint: disable=invalid-name
+        self,
+        X: Any,  # pylint: disable=invalid-name
+        **kwargs: Any,
     ) -> list[SHAPFeatureExplanation | SHAPFeatureAndAtomExplanation]:
         """Explain the predictions for the input data.
 
@@ -282,7 +284,6 @@ def explain(
             SHAPFeatureExplanation | SHAPFeatureAndAtomExplanation
         ] = []
         for input_sample in X:
-
             input_sample = [input_sample]
 
             # get predictions

molpipeline/experimental/explainability/visualization/visualization.py

@@ -358,7 +358,7 @@ def _add_shap_present_absent_features_text(
     )
 
     delta = 0.05
-    offset = offset + 0.0165
+    offset += 0.0165
     fig.text(offset, 0.13, "prediction =", ha="center", fontsize=10)
     fig.text(
         offset + 2 * delta,
@@ -558,7 +558,6 @@ def structure_heatmap_shap(  # pylint: disable=too-many-locals
     )
 
     with plt.ioff():
-
         drawer, _, _, normalizer, color_map = _structure_heatmap(
             explanation.molecule,
             explanation.atom_weights,

molpipeline/metrics/ignore_error_scorer.py

@@ -9,7 +9,7 @@
 import pandas as pd
 from loguru import logger
 from sklearn import metrics
-from sklearn.metrics._scorer import _BaseScorer  # pylint: disable=protected-access
+from sklearn.metrics._scorer import _BaseScorer
 
 
 def ignored_value_scorer(
@@ -80,10 +80,10 @@ def newscore(
             )
         y_true_ = np.copy(np.array(y_true)[all_retained])
         y_pred_ = np.array(np.array(y_pred)[all_retained].tolist())
-        _kwargs = dict(kwargs)
-        if "sample_weight" in _kwargs and _kwargs["sample_weight"] is not None:
-            _kwargs["sample_weight"] = _kwargs["sample_weight"][all_retained]
-        return score_func(y_true_, y_pred_, **_kwargs)
+        kwargs_ = dict(kwargs)
+        if "sample_weight" in kwargs_ and kwargs_["sample_weight"] is not None:
+            kwargs_["sample_weight"] = kwargs_["sample_weight"][all_retained]
+        return score_func(y_true_, y_pred_, **kwargs_)
 
     return metrics.make_scorer(
         newscore, response_method=response_method, **scorer_kwargs

molpipeline/mol2any/mol2chemprop.py

@@ -22,6 +22,7 @@
     warnings.warn(
         "chemprop not installed. MolToChemprop will not work.",
         ImportWarning,
+        stacklevel=2,
     )
 
 

molpipeline/mol2any/mol2rdkit_phys_chem.py

@@ -32,7 +32,7 @@
 # MolWt is removed as ExactMolWt is already included.
 # Ipc is removed because it causes trouble with numpy.
 DEFAULT_DESCRIPTORS = [
-    name for name in RDKIT_DESCRIPTOR_DICT if name not in ["MolWt", "Ipc"]
+    name for name in RDKIT_DESCRIPTOR_DICT if name not in {"MolWt", "Ipc"}
 ]
 
 

molpipeline/mol2mol/reaction.py

@@ -157,7 +157,9 @@ def pretransform_single(self, value: RDKitMol) -> OptionalMol:
         if len(product_list) > 1:
             if self.handle_multi == "warn":
                 warnings.warn(
-                    "Not able to handle multiple reactions. An arbitrary reaction is selected."
+                    "Not able to handle multiple reactions. "
+                    "An arbitrary reaction is selected.",
+                    stacklevel=2,
                 )
             elif self.handle_multi == "raise":
                 if mol.HasProp("identifier"):

molpipeline/utils/json_operations.py

@@ -266,7 +266,9 @@ def recursive_to_json(obj: Any) -> Any:
         # If the object is not a sklearn model, a warning is raised
         # as it might not be possible to recreate the object.
         warnings.warn(
-            f"{type(obj)} has no get_params method. No parameters for initialization are retained."
+            f"{type(obj)} has no get_params method. "
+            f"No parameters for initialization are retained.",
+            stacklevel=2,
         )
 
     return object_dict

molpipeline/utils/multi_proc.py

@@ -25,13 +25,15 @@ def check_available_cores(n_requested_cores: int) -> int:
         n_available_cores = multiprocessing.cpu_count()
     except ModuleNotFoundError:
         warnings.warn(
-            "Cannot import multiprocessing library. Falling back to single core!"
+            "Cannot import multiprocessing library. Falling back to single core!",
+            stacklevel=2,
         )
         return 1
 
     if n_requested_cores > n_available_cores:
         warnings.warn(
-            "Requested more cores than available. Using maximum number of cores!"
+            "Requested more cores than available. Using maximum number of cores!",
+            stacklevel=2,
         )
         return n_available_cores
     if n_requested_cores < 0:

notebooks/02_scaffold_split_with_custom_estimators.ipynb

@@ -403,12 +403,9 @@
     "\n",
     "    # setup the pipeline\n",
     "    pipeline = Pipeline(\n",
-    "        [(\"auto2mol\", AutoToMol())]\n",
-    "        + [\n",
-    "            (\n",
-    "                \"morgan\",\n",
-    "                MolToMorganFP(n_bits=1024, radius=2),\n",
-    "            ),\n",
+    "        [\n",
+    "            (\"auto2mol\", AutoToMol()),\n",
+    "            (\"morgan\", MolToMorganFP(n_bits=1024, radius=2)),\n",
     "            (\"RandomForestClassifier\", RandomForestClassifier(random_state=67056)),\n",
     "        ]\n",
     "    )\n",

notebooks/advanced_01_hyperopt_on_bbbp.ipynb

@@ -786,7 +786,7 @@
     "\n",
     "# Let's pre-compute the molecular standardization\n",
     "pipeline_standardization = Pipeline(\n",
-    "    [(\"auto2mol\", AutoToMol())] + make_standardization_elements(),\n",
+    "    [(\"auto2mol\", AutoToMol()), *make_standardization_elements()],\n",
     "    n_jobs=N_JOBS,\n",
     ")\n",
     "\n",
@@ -897,9 +897,11 @@
     "for model_name, model_dict in estimator_dict.items():\n",
     "\n",
     "    pipeline_estimator = Pipeline(\n",
-    "        [(\"auto2mol\", AutoToMol())]\n",
-    "        + [(\"morgan\", MolToMorganFP(return_as=\"dense\"))]\n",
-    "        + [(\"estimator\", model_dict[\"model\"])],\n",
+    "        [\n",
+    "            (\"auto2mol\", AutoToMol()),\n",
+    "            (\"morgan\", MolToMorganFP(return_as=\"dense\")),\n",
+    "            (\"estimator\", model_dict[\"model\"]),\n",
+    "        ],\n",
     "        n_jobs=N_JOBS,\n",
     "    )\n",
     "\n",
@@ -1021,7 +1023,7 @@
    "source": [
     "# pre-compute standardized molecules\n",
     "pipeline_standardization = Pipeline(\n",
-    "    [(\"auto2mol\", AutoToMol())] + make_standardization_elements(),\n",
+    "    [(\"auto2mol\", AutoToMol()), *make_standardization_elements()],\n",
     "    n_jobs=N_JOBS,\n",
     ")\n",
     "test_smiles_standardized = pipeline_standardization.transform(df_test[\"smiles\"])"

notebooks/advanced_02_add_custom_pipeline_elements.ipynb

@@ -195,7 +195,7 @@
     "\n",
     "\n",
     "class SimplifiedLogisticRegression(BaseEstimator, ClassifierMixin):\n",
-    "    \"\"\"Example estimator for the simplified logistic regression algorithm\"\"\"\n",
+    "    \"\"\"Example estimator for the simplified logistic regression algorithm.\"\"\"\n",
     "\n",
     "    def __init__(self, lr=0.01, num_iter=10000):\n",
     "        self.lr = lr\n",

notebooks/advanced_03_introduction_to_explainable_ai.ipynb

@@ -1078,12 +1078,12 @@
     "# series_dict = {k:explanations_dict[k] for k in series_names}\n",
     "series_dict = explanations_dict\n",
     "\n",
-    "weight_min = min([exp.atom_weights.min() for exp in series_dict.values()])\n",
-    "weight_max = max([exp.atom_weights.max() for exp in series_dict.values()])\n",
+    "weight_min = min(exp.atom_weights.min() for exp in series_dict.values())\n",
+    "weight_max = max(exp.atom_weights.max() for exp in series_dict.values())\n",
     "weight_abs_max = max(abs(weight_min), abs(weight_max))\n",
     "# following Harren et al., we set the maximum color intensity to 70% of the maximal numerical value\n",
     "# for better visual differentiations in low value regions.\n",
-    "weight_abs_max = weight_abs_max * 0.70"
+    "weight_abs_max *= 0.70"
    ]
   },
   {