File size: 10,866 Bytes
c49b21b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 |
#!/usr/bin/env python3
"""
Final Null Handler Integration Script
Integrates the final null value handler into the existing merge pipeline.
"""
import sys
import subprocess
from pathlib import Path
import numpy as np
import pandas as pd
from final_null_handler import FinalNullValueHandler, process_crypto_features_file, process_stock_features_file
def run_final_null_handling():
"""Run the final null value handling on all feature files"""
print("="*60)
print("STARTING FINAL NULL VALUE HANDLING")
print("="*60)
base_path = Path("data/merged/features")
files_to_process = [
("crypto_features.parquet", "crypto"),
("stocks_features.parquet", "stock"),
("merged_features.parquet", "merged")
]
results = {}
for filename, file_type in files_to_process:
file_path = base_path / filename
if not file_path.exists():
print(f"[WARNING] {filename} not found, skipping...")
continue
print(f"\n[INFO] Processing {filename}...")
try:
if file_type == "crypto":
df_processed, report = process_crypto_features_file(file_path)
elif file_type == "stock":
df_processed, report = process_stock_features_file(file_path)
elif file_type == "merged":
# For merged file, determine type by content
df_processed, report = process_merged_features_file(file_path)
results[file_type] = {
'success': True,
'file_path': file_path,
'report': report,
'rows': len(df_processed),
'nulls_filled': report['total_nulls_filled']
}
print(f"[SUCCESS] {filename} processed successfully!")
print(f" - Rows: {len(df_processed):,}")
print(f" - Nulls filled: {report['total_nulls_filled']:,}")
except Exception as e:
print(f"[ERROR] Error processing {filename}: {str(e)}")
results[file_type] = {
'success': False,
'error': str(e),
'file_path': file_path
}
return results
def process_merged_features_file(file_path):
"""Process merged features file (contains both crypto and stock data)"""
print(f"Loading merged features from {file_path}...")
df = pd.read_parquet(file_path)
print(f"Loaded {len(df)} rows with {len(df.columns)} columns")
print(f"Null values before processing: {df.isnull().sum().sum()}")
handler = FinalNullValueHandler()
# Separate crypto and stock data if possible
if 'symbol' in df.columns:
# Detect crypto vs stock based on available columns
crypto_indicators = ['rank', 'dominance', 'performance.day', 'exchangePrices.binance']
stock_indicators = ['news_activity_score_x', 'strongBuy', 'marketCapitalization']
has_crypto_cols = any(col in df.columns for col in crypto_indicators)
has_stock_cols = any(col in df.columns for col in stock_indicators)
if has_crypto_cols and has_stock_cols:
# Mixed data - process intelligently
print("Detected mixed crypto/stock data - processing intelligently...")
# Try to separate by symbol patterns or available data
crypto_mask = df['rank'].notna() | df['dominance'].notna()
if crypto_mask.any():
print(f"Processing {crypto_mask.sum()} rows as crypto data...")
df_crypto = df[crypto_mask].copy()
df_crypto_processed = handler.process_crypto_features(df_crypto)
df.loc[crypto_mask] = df_crypto_processed
stock_mask = ~crypto_mask
if stock_mask.any():
print(f"Processing {stock_mask.sum()} rows as stock data...")
df_stock = df[stock_mask].copy()
df_stock_processed = handler.process_stock_features(df_stock)
df.loc[stock_mask] = df_stock_processed
df_processed = df
elif has_crypto_cols:
print("Detected crypto-only data...")
df_processed = handler.process_crypto_features(df)
elif has_stock_cols:
print("Detected stock-only data...")
df_processed = handler.process_stock_features(df)
else:
print("Could not determine data type, applying generic processing...")
df_processed = handler.process_stock_features(df) # Default to stock processing
else:
print("No symbol column found, applying generic processing...")
df_processed = handler.process_stock_features(df)
print(f"Null values after processing: {df_processed.isnull().sum().sum()}")
# Generate report
report = handler.generate_report(df, df_processed, 'merged')
# Save processed data
df_processed.to_parquet(file_path, index=False)
print(f"Saved processed merged features to {file_path}")
return df_processed, report
def validate_data_quality(results):
"""Validate that the data quality is maintained after null handling"""
print("\n" + "="*60)
print("DATA QUALITY VALIDATION")
print("="*60)
validation_results = {}
for file_type, result in results.items():
if not result.get('success', False):
continue
file_path = result['file_path']
try:
df = pd.read_parquet(file_path)
# Basic validation checks
validation = {
'total_rows': len(df),
'total_columns': len(df.columns),
'remaining_nulls': df.isnull().sum().sum(),
'duplicate_rows': df.duplicated().sum(),
'infinite_values': np.isinf(df.select_dtypes(include=[np.number])).sum().sum(),
'data_types_consistent': True, # Could add more sophisticated checks
}
# Check for unrealistic values
numeric_cols = df.select_dtypes(include=[np.number]).columns
extreme_values = {}
for col in numeric_cols:
if col in df.columns:
col_data = df[col].dropna()
if len(col_data) > 0:
q1, q99 = col_data.quantile([0.01, 0.99])
extreme_count = ((col_data < q1 - 10 * (q99 - q1)) |
(col_data > q99 + 10 * (q99 - q1))).sum()
if extreme_count > 0:
extreme_values[col] = extreme_count
validation['extreme_values'] = extreme_values
validation['quality_score'] = calculate_quality_score(validation)
validation_results[file_type] = validation
print(f"\n{file_type.upper()} VALIDATION:")
print(f" ✓ Rows: {validation['total_rows']:,}")
print(f" ✓ Columns: {validation['total_columns']}")
print(f" ✓ Remaining nulls: {validation['remaining_nulls']}")
print(f" ✓ Duplicate rows: {validation['duplicate_rows']}")
print(f" ✓ Infinite values: {validation['infinite_values']}")
print(f" ✓ Quality score: {validation['quality_score']:.2%}")
if extreme_values:
print(f" [WARNING] Extreme values detected in {len(extreme_values)} columns")
except Exception as e:
print(f"[ERROR] Validation failed for {file_type}: {str(e)}")
validation_results[file_type] = {'error': str(e)}
return validation_results
def calculate_quality_score(validation):
"""Calculate a simple quality score"""
score = 1.0
# Penalize remaining nulls
if validation['total_rows'] > 0:
null_ratio = validation['remaining_nulls'] / (validation['total_rows'] * validation['total_columns'])
score -= null_ratio * 0.5
# Penalize duplicates
if validation['total_rows'] > 0:
dup_ratio = validation['duplicate_rows'] / validation['total_rows']
score -= dup_ratio * 0.3
# Penalize infinite values
if validation['infinite_values'] > 0:
score -= 0.1
# Penalize extreme values
extreme_columns = len(validation.get('extreme_values', {}))
if extreme_columns > 0:
score -= (extreme_columns / validation['total_columns']) * 0.2
return max(0.0, score)
def print_final_summary(results, validation_results):
"""Print final summary of the null handling process"""
print("\n" + "="*60)
print("FINAL NULL HANDLING SUMMARY")
print("="*60)
total_nulls_filled = sum(r.get('nulls_filled', 0) for r in results.values() if r.get('success'))
successful_files = sum(1 for r in results.values() if r.get('success'))
total_files = len(results)
print(f"\n[INFO] PROCESSING RESULTS:")
print(f" Files processed: {successful_files}/{total_files}")
print(f" Total nulls filled: {total_nulls_filled:,}")
print(f"\n[METRICS] QUALITY METRICS:")
for file_type, validation in validation_results.items():
if 'error' not in validation:
print(f" {file_type}: {validation['quality_score']:.1%} quality score")
if successful_files == total_files:
print(f"\n[SUCCESS] ALL FILES PROCESSED SUCCESSFULLY!")
else:
failed_files = total_files - successful_files
print(f"\n[WARNING] {failed_files} files failed to process")
print("\n[TIPS] RECOMMENDATIONS:")
print(" - Review any remaining null columns in the reports")
print(" - Monitor data quality scores in production")
print(" - Consider additional validation rules if needed")
print("\n" + "="*60)
def main():
"""Main function"""
try:
# Import numpy for validation
import numpy as np
globals()['np'] = np
# Run the null handling process
results = run_final_null_handling()
# Validate data quality
validation_results = validate_data_quality(results)
# Print final summary
print_final_summary(results, validation_results)
# Return success if all files processed successfully
success_count = sum(1 for r in results.values() if r.get('success'))
return 0 if success_count == len(results) else 1
except Exception as e:
print(f"[ERROR] Fatal error in null handling process: {str(e)}")
return 1
if __name__ == "__main__":
exit_code = main()
sys.exit(exit_code)
|