Simplify validation script - only test HuggingFace Hub loading

- Removed file structure validation
- Removed JSONL format validation
- Removed data content validation
- Removed data consistency checks
- Keep only HuggingFace Hub loading test
- Much cleaner and focused validation

validate.py

@@ -1,172 +1,15 @@
 """
-Validation script to verify the UTS2017_Bank dataset is properly created and loadable.
+Validation script to verify the UTS2017_Bank dataset can be loaded from HuggingFace Hub.
 """
 
-import json
-from pathlib import Path
-
 from datasets import load_dataset
 
 
-def validate_file_structure():
-    """Validate that all required files and directories exist."""
-    print("=" * 60)
-    print("VALIDATING FILE STRUCTURE")
-    print("=" * 60)
-
-    required_files = [
-        "data/classification/train.jsonl",
-        "data/classification/test.jsonl",
-        "data/sentiment/train.jsonl",
-        "data/sentiment/test.jsonl",
-        "data/aspect_sentiment/train.jsonl",
-        "data/aspect_sentiment/test.jsonl",
-        "raw_data/train.txt",
-        "raw_data/test.txt",
-        "README.md",
-        "preprocess.py",
-        "stats.py"
-    ]
-
-    missing_files = []
-    for file_path in required_files:
-        if not Path(file_path).exists():
-            missing_files.append(file_path)
-        else:
-            print(f"✓ {file_path}")
-
-    if missing_files:
-        print(f"\n❌ Missing files: {missing_files}")
-        return False
-    else:
-        print("\n✅ All required files present")
-        return True
-
-
-def validate_jsonl_format(file_path):
-    """Validate JSONL file format and return basic stats."""
-    try:
-        with open(file_path, encoding="utf-8") as f:
-            items = []
-            for line_num, line in enumerate(f, 1):
-                try:
-                    item = json.loads(line.strip())
-                    items.append(item)
-                except json.JSONDecodeError as e:
-                    print(f"❌ JSON error in {file_path} line {line_num}: {e}")
-                    return None
-
-        return items
-    except Exception as e:
-        print(f"❌ Error reading {file_path}: {e}")
-        return None
-
-
-def validate_data_content():
-    """Validate the content and format of each dataset subset."""
-    print("\n" + "=" * 60)
-    print("VALIDATING DATA CONTENT")
-    print("=" * 60)
-
-    # Classification subset validation
-    print("\n📊 Classification Subset:")
-    for split in ["train", "test"]:
-        file_path = f"data/classification/{split}.jsonl"
-        items = validate_jsonl_format(file_path)
-
-        if items is None:
-            continue
-
-        print(f"  {split}: {len(items)} examples")
-
-        # Check required fields
-        if items:
-            required_fields = ["text", "label"]
-            sample = items[0]
-            missing_fields = [f for f in required_fields if f not in sample]
-
-            if missing_fields:
-                print(f"    ❌ Missing fields: {missing_fields}")
-            else:
-                print("    ✓ Required fields present")
-
-            # Check label variety
-            labels = {item["label"] for item in items}
-            print(f"    ✓ {len(labels)} unique labels")
-
-    # Sentiment subset validation
-    print("\n😊 Sentiment Subset:")
-    for split in ["train", "test"]:
-        file_path = f"data/sentiment/{split}.jsonl"
-        items = validate_jsonl_format(file_path)
-
-        if items is None:
-            continue
-
-        print(f"  {split}: {len(items)} examples")
-
-        # Check required fields
-        if items:
-            required_fields = ["text", "sentiment"]
-            sample = items[0]
-            missing_fields = [f for f in required_fields if f not in sample]
-
-            if missing_fields:
-                print(f"    ❌ Missing fields: {missing_fields}")
-            else:
-                print("    ✓ Required fields present")
-
-            # Check sentiment values
-            sentiments = {item["sentiment"] for item in items}
-            expected_sentiments = {"positive", "negative", "neutral"}
-            if sentiments.issubset(expected_sentiments):
-                print(f"    ✓ Valid sentiments: {sentiments}")
-            else:
-                print(f"    ❌ Unexpected sentiments: {sentiments - expected_sentiments}")
-
-    # Aspect-sentiment subset validation
-    print("\n🎯 Aspect-Sentiment Subset:")
-    for split in ["train", "test"]:
-        file_path = f"data/aspect_sentiment/{split}.jsonl"
-        items = validate_jsonl_format(file_path)
-
-        if items is None:
-            continue
-
-        print(f"  {split}: {len(items)} examples")
-
-        # Check required fields
-        if items:
-            required_fields = ["text", "aspects"]
-            sample = items[0]
-            missing_fields = [f for f in required_fields if f not in sample]
-
-            if missing_fields:
-                print(f"    ❌ Missing fields: {missing_fields}")
-            else:
-                print("    ✓ Required fields present")
-
-            # Check aspects structure
-            if "aspects" in sample and isinstance(sample["aspects"], list):
-                if sample["aspects"]:
-                    aspect_sample = sample["aspects"][0]
-                    aspect_fields = ["aspect", "sentiment"]
-                    missing_aspect_fields = [f for f in aspect_fields if f not in aspect_sample]
-
-                    if missing_aspect_fields:
-                        print(f"    ❌ Missing aspect fields: {missing_aspect_fields}")
-                    else:
-                        print("    ✓ Aspect structure valid")
-
-                        # Count multi-aspect examples
-                        multi_aspect = sum(1 for item in items if len(item["aspects"]) > 1)
-                        print(f"    ✓ Multi-aspect examples: {multi_aspect}/{len(items)}")
-
-
 def validate_huggingface_loading():
     """Validate that the dataset can be loaded from HuggingFace Hub."""
 
-    print("\n" + "=" * 60)
+    print("🔍 UTS2017_Bank Dataset Validation")
+    print("=" * 60)
     print("VALIDATING HUGGINGFACE HUB LOADING")
     print("=" * 60)
 
@@ -187,111 +30,12 @@ def validate_huggingface_loading():
         except Exception as e:
             print(f"  ❌ Failed to load {config} from Hub: {e}")
 
-
-def validate_data_consistency():
-    """Validate data consistency across subsets."""
-    print("\n" + "=" * 60)
-    print("VALIDATING DATA CONSISTENCY")
-    print("=" * 60)
-
-    # Load all classification data to get baseline
-    classification_train = validate_jsonl_format("data/classification/train.jsonl")
-    classification_test = validate_jsonl_format("data/classification/test.jsonl")
-
-    sentiment_train = validate_jsonl_format("data/sentiment/train.jsonl")
-    sentiment_test = validate_jsonl_format("data/sentiment/test.jsonl")
-
-    aspect_train = validate_jsonl_format("data/aspect_sentiment/train.jsonl")
-    aspect_test = validate_jsonl_format("data/aspect_sentiment/test.jsonl")
-
-    if all([classification_train, classification_test, sentiment_train,
-            sentiment_test, aspect_train, aspect_test]):
-
-        # Check example counts consistency
-        train_counts = [len(classification_train), len(sentiment_train), len(aspect_train)]
-        test_counts = [len(classification_test), len(sentiment_test), len(aspect_test)]
-
-        if len(set(train_counts)) == 1:
-            print(f"✓ Train counts consistent: {train_counts[0]} examples")
-        else:
-            print(f"❌ Train counts inconsistent: {train_counts}")
-
-        if len(set(test_counts)) == 1:
-            print(f"✓ Test counts consistent: {test_counts[0]} examples")
-        else:
-            print(f"❌ Test counts inconsistent: {test_counts}")
-
-        # Check text consistency (first few examples)
-        print("\n🔍 Checking text consistency across subsets:")
-        for i in range(min(3, len(classification_train))):
-            clf_text = classification_train[i]["text"]
-            sent_text = sentiment_train[i]["text"]
-            asp_text = aspect_train[i]["text"]
-
-            if clf_text == sent_text == asp_text:
-                print(f"  ✓ Example {i+1}: Text consistent")
-            else:
-                print(f"  ❌ Example {i+1}: Text inconsistent")
-
-
-def generate_validation_summary():
-    """Generate a summary of dataset statistics."""
     print("\n" + "=" * 60)
-    print("DATASET SUMMARY")
+    print("✅ VALIDATION COMPLETE")
     print("=" * 60)
-
-    # Count total examples
-    total_train = 0
-    total_test = 0
-
-    classification_train = validate_jsonl_format("data/classification/train.jsonl")
-    classification_test = validate_jsonl_format("data/classification/test.jsonl")
-
-    if classification_train and classification_test:
-        total_train = len(classification_train)
-        total_test = len(classification_test)
-
-        print(f"📊 Total Examples: {total_train + total_test}")
-        print(f"  - Train: {total_train}")
-        print(f"  - Test: {total_test}")
-        print(f"  - Ratio: {total_train/total_test:.2f}:1")
-
-        # Count unique labels
-        labels = {item["label"] for item in classification_train + classification_test}
-        print(f"🏷️  Unique Labels: {len(labels)}")
-
-        # Most common labels
-        from collections import Counter
-        label_counts = Counter(item["label"] for item in classification_train + classification_test)
-        print("📈 Top 3 Labels:")
-        for label, count in label_counts.most_common(3):
-            print(f"  - {label}: {count}")
-
-    print("\n📁 Available Configurations:")
-    print("  - classification: Banking aspect classification")
-    print("  - sentiment: Sentiment analysis")
-    print("  - aspect_sentiment: Aspect-based sentiment analysis")
+    print("🎉 Dataset can be loaded from HuggingFace Hub!")
+    print("💡 Usage: load_dataset('undertheseanlp/UTS2017_Bank', config_name)")
 
 
 if __name__ == "__main__":
-    print("🔍 UTS2017_Bank Dataset Validation")
-    print("=" * 60)
-
-    # Run all validations
-    file_structure_ok = validate_file_structure()
-
-    if file_structure_ok:
-        validate_data_content()
-        validate_data_consistency()
-        validate_huggingface_loading()
-        generate_validation_summary()
-
-        print("\n" + "=" * 60)
-        print("✅ VALIDATION COMPLETE")
-        print("=" * 60)
-        print("🎉 Dataset appears to be properly created and formatted!")
-        print("💡 You can now use: load_dataset('undertheseanlp/UTS2017_Bank', config_name)")
-
-    else:
-        print("\n❌ VALIDATION FAILED")
-        print("Please check the missing files and run preprocessing again.")
+    validate_huggingface_loading()