From 76b8afa59cf33a9b577556874221e840b343063a Mon Sep 17 00:00:00 2001
From: Daniel Bin Schmid <danielbinschmid@outlook.de>
Date: Mon, 24 Jun 2024 21:13:42 +0200
Subject: [PATCH] fixed NamedMetrics is not a Module bug

---
 metrics/metrics.py | 53 ++++++++++++++++++++++------------------------
 1 file changed, 25 insertions(+), 28 deletions(-)

diff --git a/metrics/metrics.py b/metrics/metrics.py
index 5a37d08..1964e91 100644
--- a/metrics/metrics.py
+++ b/metrics/metrics.py
@@ -17,9 +17,9 @@ def __init__(self, metric: Metric, name: str) -> None:
 
 
 class BettiNumbersMetricCollection:
-    betti_0: ModuleList
-    betti_1: ModuleList
-    betti_2: ModuleList
+    betti_0: List[NamedMetric]
+    betti_1: List[NamedMetric]
+    betti_2: List[NamedMetric]
 
     def __init__(
         self,
@@ -127,33 +127,30 @@ def get_name_metrics(num_classes=5):
 
 def get_betti_numbers_metrics():
 
-    betti_0_metrics = ModuleList([GeneralAccuracy()])
-    betti_1_metrics = ModuleList(
-        [
-            NamedMetric(GeneralAccuracy(), "Accuracy"),
-            NamedMetric(
-                torchmetrics.classification.MulticlassAccuracy(
-                    num_classes=7,
-                    average="macro",
-                ),
-                "BalancedAccuracy",
+    betti_0_metrics = [NamedMetric(GeneralAccuracy(), "Accuracy")]
+    betti_1_metrics = [
+        NamedMetric(GeneralAccuracy(), "Accuracy"),
+        NamedMetric(
+            torchmetrics.classification.MulticlassAccuracy(
+                num_classes=7,
+                average="macro",
             ),
-        ]
-    )
-    betti_2_metrics = ModuleList(
-        [
-            NamedMetric(GeneralAccuracy(), "Accuracy"),
-            NamedMetric(MatthewsCorrCoeff(), "MCC"),
-            NamedMetric(torchmetrics.classification.BinaryF1Score(), "F1"),
-            NamedMetric(
-                torchmetrics.classification.MulticlassAccuracy(
-                    num_classes=2,
-                    average="macro",
-                ),
-                "BalancedAccuracy",
+            "BalancedAccuracy",
+        ),
+    ]
+
+    betti_2_metrics = [
+        NamedMetric(GeneralAccuracy(), "Accuracy"),
+        NamedMetric(MatthewsCorrCoeff(), "MCC"),
+        NamedMetric(torchmetrics.classification.BinaryF1Score(), "F1"),
+        NamedMetric(
+            torchmetrics.classification.MulticlassAccuracy(
+                num_classes=2,
+                average="macro",
             ),
-        ]
-    )
+            "BalancedAccuracy",
+        ),
+    ]
 
     collection = BettiNumbersMetricCollection(
         betti_0=betti_0_metrics,