Build(deps-dev): Bump ruff from 0.4.2 to 0.4.5

ethicml/implementations/adv_debiasing_method.py

@@ -55,8 +55,8 @@ def train_and_predict(
 
 def main() -> None:
     """Run the adversarial debiasing model as a standalone program."""
-    in_algo_args: "InAlgoArgs" = json.loads(sys.argv[1])
-    flags: "AdvDebArgs" = json.loads(sys.argv[2])
+    in_algo_args: InAlgoArgs = json.loads(sys.argv[1])
+    flags: AdvDebArgs = json.loads(sys.argv[2])
 
     if in_algo_args["mode"] == "run":
         train = DataTuple.from_file(Path(in_algo_args["train"]))

ethicml/implementations/agarwal.py

@@ -59,7 +59,7 @@ def fit(train: DataTuple, args: "AgarwalArgs", seed: int = 888) -> "Exponentiate
     else:
         fairness_class = EqualizedOdds(difference_bound=args["eps"])
 
-    model: "LinearModel | GradientBoostingClassifier"
+    model: LinearModel | GradientBoostingClassifier
     match classifier_type:
         case ClassifierType.svm:
             assert kernel_type is not None
@@ -119,8 +119,8 @@ def working_dir(root: Path) -> Generator[None, None, None]:
 
 def main() -> None:
     """Run the Agarwal model as a standalone program."""
-    in_algo_args: "InAlgoArgs" = json.loads(sys.argv[1])
-    flags: "AgarwalArgs" = json.loads(sys.argv[2])
+    in_algo_args: InAlgoArgs = json.loads(sys.argv[1])
+    flags: AgarwalArgs = json.loads(sys.argv[2])
     try:
         import cloudpickle  # pyright: ignore
 

ethicml/implementations/beutel.py

@@ -377,8 +377,8 @@ def _grad_reverse(features: Tensor, lambda_: float) -> Tensor:
 
 def main() -> None:
     """Load data from feather files, pass it to `train_and_transform` and then save the result."""
-    pre_algo_args: "PreAlgoArgs" = json.loads(sys.argv[1])
-    flags: "BeutelArgs" = json.loads(sys.argv[2])
+    pre_algo_args: PreAlgoArgs = json.loads(sys.argv[1])
+    flags: BeutelArgs = json.loads(sys.argv[2])
     if pre_algo_args["mode"] == "run":
         train, test = load_data_from_flags(pre_algo_args)
         save_transformations(

ethicml/implementations/dro_tabular.py

@@ -133,8 +133,8 @@ def train_and_predict(
 
 def main() -> None:
     """Run the FWD model as a standalone program on tabular data."""
-    in_algo_args: "InAlgoArgs" = json.loads(sys.argv[1])
-    flags: "DroArgs" = json.loads(sys.argv[2])
+    in_algo_args: InAlgoArgs = json.loads(sys.argv[1])
+    flags: DroArgs = json.loads(sys.argv[2])
     data: DataTuple | TestTuple
     if in_algo_args["mode"] == "run":
         train, test = load_data_from_flags(in_algo_args)

ethicml/implementations/fair_dummies_romano.py

@@ -59,8 +59,8 @@ def train_and_predict(
 
 def main() -> None:
     """Run the Agarwal model as a standalone program."""
-    in_algo_args: "InAlgoArgs" = json.loads(sys.argv[1])
-    flags: "FairDummiesArgs" = json.loads(sys.argv[2])
+    in_algo_args: InAlgoArgs = json.loads(sys.argv[1])
+    flags: FairDummiesArgs = json.loads(sys.argv[2])
 
     if in_algo_args["mode"] == "run":
         train = DataTuple.from_file(Path(in_algo_args["train"]))

ethicml/implementations/hgr_method.py

@@ -61,8 +61,8 @@ def train_and_predict(
 
 def main() -> None:
     """Run the Agarwal model as a standalone program."""
-    in_algo_args: "InAlgoArgs" = json.loads(sys.argv[1])
-    flags: "HgrArgs" = json.loads(sys.argv[2])
+    in_algo_args: InAlgoArgs = json.loads(sys.argv[1])
+    flags: HgrArgs = json.loads(sys.argv[2])
 
     if in_algo_args["mode"] == "run":
         train = DataTuple.from_file(Path(in_algo_args["train"]))

ethicml/implementations/vfae.py

@@ -136,8 +136,8 @@ def train_model(
 
 def main() -> None:
     """Run model."""
-    pre_algo_args: "PreAlgoArgs" = json.loads(sys.argv[1])
-    flags: "VfaeArgs" = json.loads(sys.argv[2])
+    pre_algo_args: PreAlgoArgs = json.loads(sys.argv[1])
+    flags: VfaeArgs = json.loads(sys.argv[2])
     if pre_algo_args["mode"] == "run":
         set_seed(pre_algo_args["seed"])
         train, test = load_data_from_flags(pre_algo_args)

ethicml/implementations/zemel.py

@@ -196,8 +196,8 @@ def main() -> None:
     Based on code from https://github.com/zjelveh/learning-fair-representations
     Which in turn, we've got from AIF360
     """
-    pre_algo_args: "PreAlgoArgs" = json.loads(sys.argv[1])
-    flags: "ZemelArgs" = json.loads(sys.argv[2])
+    pre_algo_args: PreAlgoArgs = json.loads(sys.argv[1])
+    flags: ZemelArgs = json.loads(sys.argv[2])
     if pre_algo_args["mode"] == "run":
         train, test = load_data_from_flags(pre_algo_args)
         save_transformations(