updated the degradation analysis

langtest/transform/accuracy.py

@@ -1146,14 +1146,14 @@ class DegradationAnalysis(BaseAccuracy):
 
     supported_tasks = ["ner", "text-classification"]
 
+    result_data = defaultdict(dict)
+
     @classmethod
     def transform(cls, test: str, y_true: List[Any], params: Dict):
-        sample = MinScoreSample(
-            category="accuracy",
-            test_type="degradation_analysis",
-        )
+        # reset the result data
+        DegradationAnalysis.result_data.clear()
 
-        return [sample]
+        return []
 
     @staticmethod
     async def run(
@@ -1183,8 +1183,6 @@ async def run(
 
         progress = kwargs.get("progress_bar", False)
 
-        output = defaultdict(dict)
-
         for category, data in test_cases.items():
             if category not in ["robustness", "bias"]:
                 continue
@@ -1205,24 +1203,30 @@ async def run(
 
                 degradation = accuracy_score2 - accuracy_score1
 
-                output[category][test_type] = degradation
+                DegradationAnalysis.result_data[category][test_type] = {
+                    "before": accuracy_score1,
+                    "after": accuracy_score2,
+                    "difference": degradation,
+                }
             if progress:
                 progress.update(1)
 
-        print(output)
         return []
 
     @staticmethod
-    def preprocess(y_true: List, y_pred: List):
+    def preprocess(y_true: Union[list, pd.Series], y_pred: Union[list, pd.Series]):
         """
         Preprocesses the input data for the degradation analysis.
 
         Args:
+
             y_true (List): The true labels.
             y_pred (List): The predicted labels.
 
         Returns:
-            Tuple[pd.Series, pd.Series]: The preprocessed true and predicted labels.
+
+            y_true, y_pred (Tuple[pd.Series, pd.Series]):
+            The preprocessed true and predicted labels.
         """
 
         if isinstance(y_true, list):
@@ -1246,3 +1250,72 @@ def preprocess(y_true: List, y_pred: List):
         y_true = y_true.apply(lambda x: x.split("-")[-1])
 
         return y_true, y_pred
+
+    @staticmethod
+    def show_results():
+        import pandas as pd
+        import matplotlib.pyplot as plt
+
+        data = DegradationAnalysis.result_data
+        if not data:
+            raise ValueError("No data found for degradation analysis.")
+
+        for category, tests in data.items():
+            df = pd.DataFrame(tests).T
+
+            fig, ax = plt.subplots(figsize=(12, 6))
+
+            y_labels = df.index
+            y_pos = range(len(y_labels))
+
+            for i, label in enumerate(y_labels):
+                # Before robustness bar
+                ax.broken_barh(
+                    [(0, df["after"][i])],
+                    (i - 0.2, 0.4),
+                    color="#1f77b4",
+                    label="After" if i == 0 else "",
+                )
+                # After robustness bar with adjusted width send this bar to back
+                ax.broken_barh(
+                    [(0, df["before"][i])],
+                    (i - 0.4, 0.8),
+                    color="#d3d3d3",
+                    zorder=0,
+                    label="Before" if i == 0 else "",
+                )
+
+                # Adjust label positions if too close
+                if abs(df["before"][i] - df["after"][i]) < 0.05:
+                    offset = 0.03
+                else:
+                    offset = 0.01
+
+                ax.text(
+                    df["after"][i] + 0.01,
+                    i,
+                    f"{df['after'][i]:.2f}",
+                    va="center",
+                    ha="left",
+                    color="#1f77b4",
+                )
+                ax.text(
+                    df["before"][i] + offset,
+                    i,
+                    f"{df['before'][i]:.2f}",
+                    va="center",
+                    ha="left",
+                    color="black",
+                )
+
+            ax.set_xlim(0, 1)
+            ax.set_yticks(y_pos)
+            ax.set_yticklabels(y_labels)
+            ax.set_xlabel("Accuracy Score Over Robustness and Bias Tests")
+            ax.set_title(
+                "Comparison of Accuracy Before and After Robustness and Bias Tests"
+            )
+            ax.legend()
+
+            plt.tight_layout()
+            plt.show()