formatting

code/datasets/simplicial_ds.py

@@ -103,7 +103,10 @@ def processed_file_names(self):
         else:
             f_names = [
                 self._data_filename(task_type, mode)
-                for task_type in [TaskType.BETTI_NUMBERS, TaskType.ORIENTABILITY]
+                for task_type in [
+                    TaskType.BETTI_NUMBERS,
+                    TaskType.ORIENTABILITY,
+                ]
                 for mode in ["train", "test", "val"]
             ]
         return f_names
@@ -115,9 +118,7 @@ def process(self):
         for task_type in TaskType:
 
             # no name classification on 3 manifolds
-            if self.manifold == "3" and (
-                task_type == TaskType.NAME
-            ):
+            if self.manifold == "3" and (task_type == TaskType.NAME):
                 continue
 
             # apply class transform

code/datasets/transforms.py

@@ -10,6 +10,7 @@
 from enum import Enum
 from datasets.dataset_types import DatasetType
 from typing import List, Callable
+
 NAME_TO_CLASS = {"Klein bottle": 0, "RP^2": 1, "T^2": 2, "S^2": 3, "": 4}
 
 
@@ -21,7 +22,7 @@ def __call__(self, data):
 
 class OrientableToClassTransform:
     def __call__(self, data):
-        data.orientable = torch.tensor(data.betti_numbers)[...,-1]
+        data.orientable = torch.tensor(data.betti_numbers)[..., -1]
         data.y = data.orientable.long()
         return data
 
@@ -52,6 +53,7 @@ def __call__(self, data):
         data.triangulation = None
         return data
 
+
 class TriangulationToFaceTransform:
     """
     Transforms tetrahedra to faces.
@@ -71,7 +73,9 @@ def __call__(self, data):
         idx = [[0, 1, 2], [0, 1, 3], [0, 2, 3], [1, 2, 3]]
         if hasattr(data, "triangulation"):
             assert data.triangulation is not None
-            face = torch.cat([torch.tensor(data.triangulation)[i] for i in idx], dim=1)
+            face = torch.cat(
+                [torch.tensor(data.triangulation)[i] for i in idx], dim=1
+            )
 
             # Remove duplicate triangles in
             data.face = torch.unique(face, dim=1)
@@ -81,6 +85,7 @@ def __call__(self, data):
 
         return data
 
+
 class SimplicialComplexTransform:
     def __call__(self, data):
         data.sc = SimplicialComplex(data.triangulation)
@@ -242,8 +247,9 @@ class TransformType(Enum):
     random_simplices_features = "random_simplices_features"
 
 
-
-def transforms_lookup(tr_type: TransformType, ds_type: DatasetType) -> List[Callable]:
+def transforms_lookup(
+    tr_type: TransformType, ds_type: DatasetType
+) -> List[Callable]:
     _transforms_lookup = {
         TransformType.degree_transform: degree_transform,
         TransformType.degree_transform_onehot: degree_transform_onehot,
@@ -253,7 +259,10 @@ def transforms_lookup(tr_type: TransformType, ds_type: DatasetType) -> List[Call
     }
 
     tr = _transforms_lookup[tr_type]
-    if tr_type != TransformType.degree_transform_sc and tr_type != TransformType.random_simplices_features:
+    if (
+        tr_type != TransformType.degree_transform_sc
+        and tr_type != TransformType.random_simplices_features
+    ):
         tr[0] = TriangulationToFaceTransform()
-    
+
     return tr

code/metrics/metrics.py

@@ -43,7 +43,7 @@ def __init__(
         betti_0: List[NamedMetric],
         betti_1: List[NamedMetric],
         betti_2: List[NamedMetric],
-        betti_3: Optional[List[NamedMetric]] = None
+        betti_3: Optional[List[NamedMetric]] = None,
     ) -> None:
         self.betti_0 = betti_0
         self.betti_1 = betti_1
@@ -53,13 +53,9 @@ def __init__(
     def as_list(self):
         if self.betti_3 is None:
             return [self.betti_0, self.betti_1, self.betti_2]
-        else: 
-            return [
-                self.betti_0,
-                self.betti_1,
-                self.betti_2,
-                self.betti_3
-            ]
+        else:
+            return [self.betti_0, self.betti_1, self.betti_2, self.betti_3]
+
 
 class MetricTrainValTest:
     """
@@ -154,7 +150,9 @@ def get_betti_numbers_metrics(ds_type: DatasetType):
         betti_2_metrics = [
             NamedMetric(GeneralAccuracy(), "Accuracy"),
             NamedMetric(MatthewsCorrCoeff(), "MCC"),
-            NamedMetric(torchmetrics.classification.BinaryAUROC(), "BinaryAUROC"),
+            NamedMetric(
+                torchmetrics.classification.BinaryAUROC(), "BinaryAUROC"
+            ),
             NamedMetric(
                 BettiNumbersMultiClassAccuracy(num_classes=2),
                 "BalancedAccuracy",

code/metrics/tasks.py

@@ -106,5 +106,5 @@ def get_task_lookup(
 
 class_transforms_lookup_3manifold: Dict[TaskType, List[Callable]] = {
     TaskType.BETTI_NUMBERS: betti_numbers_transforms_3manifold,
-    TaskType.ORIENTABILITY: orientability_transforms
+    TaskType.ORIENTABILITY: orientability_transforms,
 }