Roni Goldshmidt
commited on
Commit
·
436a7cd
1
Parent(s):
7ada547
Initial leaderboard setup
Browse files
.ipynb_checkpoints/comparison-checkpoint.py
CHANGED
|
@@ -42,15 +42,21 @@ class ModelEvaluator:
|
|
| 42 |
print(f"Skipping {category} - missing columns")
|
| 43 |
continue
|
| 44 |
|
| 45 |
-
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
|
| 48 |
valid_labels = sorted(set(y_true) | set(y_pred))
|
| 49 |
|
| 50 |
-
valid_labels = [label for label in valid_labels if (y_true == label).sum() > 0]
|
| 51 |
|
| 52 |
if not valid_labels:
|
| 53 |
-
print(f"Skipping {category} - No valid labels found.")
|
| 54 |
continue
|
| 55 |
|
| 56 |
class_precisions = precision_score(y_true, y_pred, labels=valid_labels, average=None, zero_division=0)
|
|
|
|
| 42 |
print(f"Skipping {category} - missing columns")
|
| 43 |
continue
|
| 44 |
|
| 45 |
+
filtered_df = merged_df[merged_df[true_col] != "unknown"]
|
| 46 |
+
|
| 47 |
+
if filtered_df.empty:
|
| 48 |
+
print(f"Skipping {category} - only 'unknown' values present.")
|
| 49 |
+
continue
|
| 50 |
+
|
| 51 |
+
y_true = filtered_df[true_col].astype(str)
|
| 52 |
+
y_pred = filtered_df[pred_col].astype(str)
|
| 53 |
|
| 54 |
valid_labels = sorted(set(y_true) | set(y_pred))
|
| 55 |
|
| 56 |
+
valid_labels = [label for label in valid_labels if (y_true == label).sum() > 0 and label != "unknown"]
|
| 57 |
|
| 58 |
if not valid_labels:
|
| 59 |
+
print(f"Skipping {category} - No valid labels found after filtering.")
|
| 60 |
continue
|
| 61 |
|
| 62 |
class_precisions = precision_score(y_true, y_pred, labels=valid_labels, average=None, zero_division=0)
|
__pycache__/comparison.cpython-310.pyc
CHANGED
|
Binary files a/__pycache__/comparison.cpython-310.pyc and b/__pycache__/comparison.cpython-310.pyc differ
|
|
|
comparison.py
CHANGED
|
@@ -42,15 +42,21 @@ class ModelEvaluator:
|
|
| 42 |
print(f"Skipping {category} - missing columns")
|
| 43 |
continue
|
| 44 |
|
| 45 |
-
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
|
| 48 |
valid_labels = sorted(set(y_true) | set(y_pred))
|
| 49 |
|
| 50 |
-
valid_labels = [label for label in valid_labels if (y_true == label).sum() > 0]
|
| 51 |
|
| 52 |
if not valid_labels:
|
| 53 |
-
print(f"Skipping {category} - No valid labels found.")
|
| 54 |
continue
|
| 55 |
|
| 56 |
class_precisions = precision_score(y_true, y_pred, labels=valid_labels, average=None, zero_division=0)
|
|
|
|
| 42 |
print(f"Skipping {category} - missing columns")
|
| 43 |
continue
|
| 44 |
|
| 45 |
+
filtered_df = merged_df[merged_df[true_col] != "unknown"]
|
| 46 |
+
|
| 47 |
+
if filtered_df.empty:
|
| 48 |
+
print(f"Skipping {category} - only 'unknown' values present.")
|
| 49 |
+
continue
|
| 50 |
+
|
| 51 |
+
y_true = filtered_df[true_col].astype(str)
|
| 52 |
+
y_pred = filtered_df[pred_col].astype(str)
|
| 53 |
|
| 54 |
valid_labels = sorted(set(y_true) | set(y_pred))
|
| 55 |
|
| 56 |
+
valid_labels = [label for label in valid_labels if (y_true == label).sum() > 0 and label != "unknown"]
|
| 57 |
|
| 58 |
if not valid_labels:
|
| 59 |
+
print(f"Skipping {category} - No valid labels found after filtering.")
|
| 60 |
continue
|
| 61 |
|
| 62 |
class_precisions = precision_score(y_true, y_pred, labels=valid_labels, average=None, zero_division=0)
|