File size: 38,029 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 |
"""
Santiment Data Merger
=====================
This script merges all Santiment data files into a unified features dataset.
It reads all parquet files from data/santiment/, merges them by slug and datetime
with 1-hour interval tolerance, and creates merged_features.parquet.
Features:
- Reads all Santiment parquet files automatically
- Merges by slug and datetime with 1-hour tolerance
- Handles different data formats (financial, ohlcv, prices, etc.)
- Creates comprehensive feature dataset
- Robust error handling and logging
Author: AI Assistant
Date: August 2025
"""
import os
import sys
import pandas as pd
import numpy as np
from pathlib import Path
from datetime import datetime, timedelta
import logging
import glob
from typing import List, Dict, Optional, Tuple
import warnings
# Resolve data directory base
try:
from src.config import DATA_DIR as CFG_DATA_DIR
except Exception:
try:
from config import DATA_DIR as CFG_DATA_DIR
except Exception:
CFG_DATA_DIR = "/data"
def _resolve_under_data(path_like: str | os.PathLike) -> Path:
p = Path(path_like)
if p.is_absolute():
return p
parts = p.parts
if parts and parts[0].lower() == "data":
rel = Path(*parts[1:]) if len(parts) > 1 else Path()
else:
rel = p
return Path(CFG_DATA_DIR) / rel
# Configure logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)
class SantimentDataMerger:
"""
Comprehensive Santiment Data Merger
Merges all Santiment parquet files into a unified features dataset
with proper handling of different data formats and time alignment.
"""
def __init__(self,
source_dir: str = "data/santiment",
output_dir: str = "data/santiment",
time_tolerance_hours: int = 1):
"""
Initialize the Santiment Data Merger
Args:
source_dir: Directory containing Santiment parquet files
output_dir: Directory to save merged features
time_tolerance_hours: Tolerance for datetime matching (hours)
"""
# Resolve under DATA_DIR for portability
self.source_dir = _resolve_under_data(source_dir)
self.output_dir = _resolve_under_data(output_dir)
self.time_tolerance = timedelta(hours=time_tolerance_hours)
# Ensure directories exist
self.source_dir.mkdir(parents=True, exist_ok=True)
self.output_dir.mkdir(parents=True, exist_ok=True)
# Storage for processed data
self.dataframes: Dict[str, pd.DataFrame] = {}
self.merged_data: Optional[pd.DataFrame] = None
self.processing_stats = {
'files_found': 0,
'files_processed': 0,
'files_failed': 0,
'total_records': 0,
'unique_slugs': set(),
'date_range': {},
'categories': set()
}
# Track placeholder mode (no input files)
self.placeholder_created = False
# Initialize symbol normalizer
self.symbol_normalizer = self._setup_symbol_normalizer()
def _setup_symbol_normalizer(self):
"""
Set up symbol normalization mapping for consistent asset identification
Returns:
Dictionary mapping various symbol formats to canonical slugs
"""
# Canonical mapping for major crypto assets
# Maps various symbols/names to the official uppercase symbols
symbol_mapping = {
# Bitcoin variants
'bitcoin': 'BTC',
'btc': 'BTC',
'Bitcoin': 'BTC',
'BTC': 'BTC',
# Ethereum variants
'ethereum': 'ETH',
'eth': 'ETH',
'Ethereum': 'ETH',
'ETH': 'ETH',
# Ripple/XRP variants
'ripple': 'XRP',
'xrp': 'XRP',
'Ripple': 'XRP',
'XRP': 'XRP',
# Solana variants
'solana': 'SOL',
'sol': 'SOL',
'Solana': 'SOL',
'SOL': 'SOL',
# Cardano variants
'cardano': 'ADA',
'ada': 'ADA',
'Cardano': 'ADA',
'ADA': 'ADA',
# Polkadot variants
'polkadot': 'DOT',
'dot': 'DOT',
'Polkadot': 'DOT',
'DOT': 'DOT',
# Chainlink variants
'chainlink': 'LINK',
'link': 'LINK',
'Chainlink': 'LINK',
'LINK': 'LINK',
# Litecoin variants
'litecoin': 'LTC',
'ltc': 'LTC',
'Litecoin': 'LTC',
'LTC': 'LTC',
# Bitcoin Cash variants
'bitcoin-cash': 'BCH',
'bch': 'BCH',
'Bitcoin Cash': 'BCH',
'BCH': 'BCH',
# Stellar variants
'stellar': 'XLM',
'xlm': 'XLM',
'Stellar': 'XLM',
'XLM': 'XLM',
# Ethereum Classic variants
'ethereum-classic': 'ETC',
'etc': 'ETC',
'Ethereum Classic': 'ETC',
'ETC': 'ETC',
# EOS variants
'eos': 'EOS',
'EOS': 'EOS',
}
logger.info(f"Initialized symbol normalizer with {len(symbol_mapping)} mappings")
return symbol_mapping
def normalize_symbol(self, symbol: str) -> str:
"""
Normalize a symbol to its canonical uppercase format
Args:
symbol: Symbol to normalize
Returns:
Canonical uppercase symbol (e.g., BTC, ETH, SOL)
"""
if symbol in self.symbol_normalizer:
canonical = self.symbol_normalizer[symbol]
if symbol != canonical:
logger.debug(f"Normalized '{symbol}' -> '{canonical}'")
return canonical
# If not found in mapping, return uppercase version and log warning
logger.warning(f"Unknown symbol '{symbol}' not found in normalization mapping, using uppercase")
return symbol.upper()
def find_parquet_files(self) -> List[Path]:
"""
Find all parquet files in the source directory
Returns:
List of parquet file paths
"""
parquet_files = list(self.source_dir.glob("*.parquet"))
# Filter out non-Santiment files and already merged files
santiment_files = []
for file_path in parquet_files:
filename = file_path.name.lower()
# Include Santiment files but exclude already merged ones
if ('santiment_' in filename or 'ohlcv' in filename or 'prices' in filename) and 'merged' not in filename:
santiment_files.append(file_path)
self.processing_stats['files_found'] = len(santiment_files)
logger.info(f"Found {len(santiment_files)} Santiment parquet files")
return santiment_files
def parse_filename(self, file_path: Path) -> Dict[str, str]:
"""
Parse filename to extract metadata
Args:
file_path: Path to the parquet file
Returns:
Dictionary with parsed metadata
"""
filename = file_path.stem
parts = filename.split('_')
metadata = {
'source': 'santiment',
'category': 'unknown',
'metric': 'unknown',
'asset': 'unknown',
'timestamp': 'unknown'
}
try:
if filename.startswith('santiment_'):
# Format: santiment_category_metric_timestamp
if len(parts) >= 4:
metadata['category'] = parts[1]
metadata['metric'] = parts[2]
metadata['timestamp'] = '_'.join(parts[3:])
elif 'ohlcv' in filename:
# Format: santiment_ohlcv_asset_timestamp
if len(parts) >= 4:
metadata['category'] = 'ohlcv'
metadata['metric'] = 'ohlcv'
metadata['asset'] = parts[2]
metadata['timestamp'] = '_'.join(parts[3:])
elif 'prices' in filename:
# Format: santiment_prices_asset_timestamp
if len(parts) >= 4:
metadata['category'] = 'prices'
metadata['metric'] = 'prices_detailed'
metadata['asset'] = parts[2]
metadata['timestamp'] = '_'.join(parts[3:])
except Exception as e:
logger.warning(f"Failed to parse filename {filename}: {e}")
return metadata
def load_and_standardize_dataframe(self, file_path: Path) -> Optional[pd.DataFrame]:
"""
Load and standardize a parquet file
Args:
file_path: Path to the parquet file
Returns:
Standardized DataFrame or None if failed
"""
try:
df = pd.read_parquet(file_path)
if df.empty:
logger.warning(f"Empty dataframe: {file_path.name}")
return None
# Parse filename for metadata
metadata = self.parse_filename(file_path)
# Standardize datetime index
if 'datetime' in df.columns:
df['datetime'] = pd.to_datetime(df['datetime'])
df.set_index('datetime', inplace=True)
elif df.index.name == 'datetime' or pd.api.types.is_datetime64_any_dtype(df.index):
df.index = pd.to_datetime(df.index)
df.index.name = 'datetime'
else:
# Try to find a datetime column
datetime_cols = [col for col in df.columns if 'date' in col.lower() or 'time' in col.lower()]
if datetime_cols:
df[datetime_cols[0]] = pd.to_datetime(df[datetime_cols[0]])
df.set_index(datetime_cols[0], inplace=True)
df.index.name = 'datetime'
else:
logger.warning(f"No datetime column found in {file_path.name}")
return None
# Ensure slug column exists
if 'slug' not in df.columns:
if metadata['asset'] != 'unknown':
# Normalize the asset symbol before assigning
normalized_asset = self.normalize_symbol(metadata['asset'])
df['slug'] = normalized_asset
if metadata['asset'] != normalized_asset:
logger.info(f"Normalized asset '{metadata['asset']}' -> '{normalized_asset}' in {file_path.name}")
else:
logger.warning(f"No slug information found in {file_path.name}")
return None
else:
# Normalize existing slug column
df['slug'] = df['slug'].apply(self.normalize_symbol)
logger.debug(f"Normalized existing slug column in {file_path.name}")
# Add metadata columns
df['source_file'] = file_path.name
df['category'] = metadata['category']
# Rename columns to avoid conflicts and add prefixes
value_columns = [col for col in df.columns if col not in ['slug', 'metric', 'source_file', 'category']]
# Add category prefix to value columns
category = metadata['category']
metric = metadata['metric']
column_mapping = {}
for col in value_columns:
if col in ['slug', 'source_file', 'category']:
continue
# Create meaningful column name
if col == 'value':
new_col = f"{category}_{metric}"
elif col in ['open', 'high', 'low', 'close', 'volume']:
new_col = f"{category}_{col}"
else:
new_col = f"{category}_{col}"
column_mapping[col] = new_col
df.rename(columns=column_mapping, inplace=True)
# Update stats
self.processing_stats['unique_slugs'].update(df['slug'].unique())
self.processing_stats['categories'].add(category)
logger.info(f"Loaded {file_path.name}: {len(df)} records, {len(df.columns)} columns")
return df
except Exception as e:
logger.error(f"Failed to load {file_path.name}: {e}")
return None
def merge_dataframes_by_slug_datetime(self, dataframes: List[pd.DataFrame]) -> pd.DataFrame:
"""
Merge multiple dataframes by slug and datetime with tolerance
Args:
dataframes: List of DataFrames to merge
Returns:
Merged DataFrame
"""
if not dataframes:
return pd.DataFrame()
logger.info(f"Merging {len(dataframes)} dataframes...")
# Start with the first dataframe
merged = dataframes[0].copy()
logger.info(f"Starting with base dataframe: {len(merged)} records")
# Merge each subsequent dataframe
for i, df in enumerate(dataframes[1:], 1):
logger.info(f"Merging dataframe {i+1}/{len(dataframes)}: {len(df)} records")
try:
# Merge on slug and datetime index with tolerance
merged = self._merge_with_time_tolerance(merged, df)
logger.info(f"After merge {i}: {len(merged)} records")
except Exception as e:
logger.error(f"Failed to merge dataframe {i+1}: {e}")
continue
return merged
def _merge_with_time_tolerance(self, left_df: pd.DataFrame, right_df: pd.DataFrame) -> pd.DataFrame:
"""
Merge two dataframes with time tolerance
Args:
left_df: Left DataFrame
right_df: Right DataFrame
Returns:
Merged DataFrame
"""
# Reset index to make datetime a column for merging
left_reset = left_df.reset_index()
right_reset = right_df.reset_index()
# Perform merge on slug first
common_slugs = set(left_reset['slug'].unique()) & set(right_reset['slug'].unique())
if not common_slugs:
# No common slugs, concatenate vertically
logger.warning("No common slugs found, concatenating dataframes")
combined = pd.concat([left_df, right_df], axis=0, sort=False)
return combined.sort_index()
merged_parts = []
for slug in common_slugs:
left_slug = left_reset[left_reset['slug'] == slug].copy()
right_slug = right_reset[right_reset['slug'] == slug].copy()
if left_slug.empty or right_slug.empty:
continue
# Sort by datetime
left_slug = left_slug.sort_values('datetime')
right_slug = right_slug.sort_values('datetime')
# Merge with time tolerance using pandas merge_asof
try:
merged_slug = pd.merge_asof(
left_slug,
right_slug,
on='datetime',
by='slug',
tolerance=self.time_tolerance,
direction='nearest',
suffixes=('', '_right')
)
# Remove duplicate columns
duplicate_cols = [col for col in merged_slug.columns if col.endswith('_right')]
for col in duplicate_cols:
base_col = col.replace('_right', '')
if base_col in merged_slug.columns:
# Keep non-null values, preferring left side
merged_slug[base_col] = merged_slug[base_col].fillna(merged_slug[col])
else:
# Rename the right column
merged_slug[base_col] = merged_slug[col]
merged_slug.drop(columns=[col], inplace=True)
merged_parts.append(merged_slug)
except Exception as e:
logger.warning(f"Failed to merge slug {slug}: {e}")
# Fallback: simple concatenation for this slug
slug_combined = pd.concat([left_slug, right_slug], axis=0, sort=False)
merged_parts.append(slug_combined)
# Handle slugs that exist in only one dataframe
left_only_slugs = set(left_reset['slug'].unique()) - common_slugs
right_only_slugs = set(right_reset['slug'].unique()) - common_slugs
for slug in left_only_slugs:
merged_parts.append(left_reset[left_reset['slug'] == slug])
for slug in right_only_slugs:
merged_parts.append(right_reset[right_reset['slug'] == slug])
# Combine all parts
if merged_parts:
final_merged = pd.concat(merged_parts, axis=0, sort=False, ignore_index=True)
# Set datetime as index
final_merged.set_index('datetime', inplace=True)
return final_merged.sort_index()
else:
return left_df
def fill_missing_values(self, df: pd.DataFrame) -> pd.DataFrame:
"""
Comprehensive null filling strategy for the merged dataset
Args:
df: DataFrame with potential null values
Returns:
DataFrame with filled null values
"""
logger.info("Applying comprehensive null filling strategy...")
filled_df = df.copy()
null_counts_before = filled_df.isnull().sum().sum()
# Strategy 1: Forward fill within each asset (time-based continuity)
logger.info("Step 1: Forward filling within each asset...")
for slug in filled_df['slug'].unique():
slug_mask = filled_df['slug'] == slug
filled_df.loc[slug_mask] = filled_df.loc[slug_mask].ffill()
# Strategy 2: Backward fill within each asset (fill initial nulls)
logger.info("Step 2: Backward filling within each asset...")
for slug in filled_df['slug'].unique():
slug_mask = filled_df['slug'] == slug
filled_df.loc[slug_mask] = filled_df.loc[slug_mask].bfill()
# Strategy 3: Fill specific column types with appropriate defaults
logger.info("Step 3: Filling remaining nulls with type-specific defaults...")
for col in filled_df.columns:
if filled_df[col].isnull().any():
# Price and financial metrics: use median of the column
if any(keyword in col.lower() for keyword in ['price', 'usd', 'btc', 'eth', 'marketcap', 'volume']):
median_val = filled_df[col].median()
filled_df[col] = filled_df[col].fillna(median_val)
logger.debug(f"Filled {col} nulls with median: {median_val}")
# Address and network metrics: use 0 (no activity)
elif any(keyword in col.lower() for keyword in ['address', 'network', 'active', 'transaction']):
filled_df[col] = filled_df[col].fillna(0)
logger.debug(f"Filled {col} nulls with 0")
# Exchange metrics: use 0 (no flow)
elif any(keyword in col.lower() for keyword in ['exchange', 'inflow', 'outflow', 'balance']):
filled_df[col] = filled_df[col].fillna(0)
logger.debug(f"Filled {col} nulls with 0")
# Supply metrics: forward fill or use mean
elif any(keyword in col.lower() for keyword in ['supply', 'circulation', 'velocity']):
mean_val = filled_df[col].mean()
filled_df[col] = filled_df[col].fillna(mean_val)
logger.debug(f"Filled {col} nulls with mean: {mean_val}")
# Development metrics: use 0 (no activity)
elif any(keyword in col.lower() for keyword in ['dev', 'github', 'contributors']):
filled_df[col] = filled_df[col].fillna(0)
logger.debug(f"Filled {col} nulls with 0")
# Social metrics: use 0 (no mentions)
elif any(keyword in col.lower() for keyword in ['social', 'sentiment', 'volume_4chan', 'volume_reddit']):
filled_df[col] = filled_df[col].fillna(0)
logger.debug(f"Filled {col} nulls with 0")
# OHLCV metrics: use forward fill or interpolation
elif any(keyword in col.lower() for keyword in ['open', 'high', 'low', 'close', 'ohlcv']):
filled_df[col] = filled_df[col].ffill().bfill()
logger.debug(f"Filled {col} nulls with forward/backward fill")
# Derivatives and whale metrics: use 0
elif any(keyword in col.lower() for keyword in ['funding', 'interest', 'whale', 'holders']):
filled_df[col] = filled_df[col].fillna(0)
logger.debug(f"Filled {col} nulls with 0")
# String columns: use 'unknown' or most frequent value
elif filled_df[col].dtype == 'object':
if col in ['slug', 'category', 'source_file', 'metric', 'development_alternative_slug_used']:
# Skip these columns as they will be removed or are handled separately
continue
else:
mode_val = filled_df[col].mode()
if len(mode_val) > 0:
filled_df[col] = filled_df[col].fillna(mode_val[0])
else:
filled_df[col] = filled_df[col].fillna('unknown')
logger.debug(f"Filled {col} nulls with mode/unknown")
# Any remaining numeric nulls: use median
elif pd.api.types.is_numeric_dtype(filled_df[col]):
median_val = filled_df[col].median()
if pd.notna(median_val):
filled_df[col] = filled_df[col].fillna(median_val)
logger.debug(f"Filled {col} nulls with median: {median_val}")
else:
filled_df[col] = filled_df[col].fillna(0)
logger.debug(f"Filled {col} nulls with 0 (median was NaN)")
null_counts_after = filled_df.isnull().sum().sum()
nulls_filled = null_counts_before - null_counts_after
logger.info(f"Null filling completed:")
logger.info(f" Nulls before: {null_counts_before:,}")
logger.info(f" Nulls after: {null_counts_after:,}")
logger.info(f" Nulls filled: {nulls_filled:,}")
return filled_df
def process_all_files(self) -> bool:
"""
Process all Santiment parquet files
Returns:
True if successful, False otherwise
"""
try:
# Find all parquet files
parquet_files = self.find_parquet_files()
if not parquet_files:
logger.warning("No Santiment parquet files found")
# Graceful fallback: create minimal placeholder merged file to unblock pipeline
try:
# Create an explicitly typed empty DF with expected columns
placeholder = pd.DataFrame({'slug': pd.Series(dtype='object')})
# Set an empty datetime index (naive) with the expected name
placeholder.index = pd.DatetimeIndex([], name='datetime')
# Ensure output directory exists
self.output_dir.mkdir(parents=True, exist_ok=True)
out_path = self.output_dir / "merged_features.parquet"
# Save directly, bypassing save_merged_features constraints
placeholder.to_parquet(out_path, index=True)
# Mark placeholder state and keep merged_data None
self.placeholder_created = True
logger.info(f"Created placeholder Santiment merged_features.parquet with 0 rows at {out_path}")
return True
except Exception as e:
logger.error(f"Failed to create placeholder Santiment file: {e}")
return False
# Load and standardize all dataframes
dataframes = []
for file_path in parquet_files:
try:
df = self.load_and_standardize_dataframe(file_path)
if df is not None:
dataframes.append(df)
self.processing_stats['files_processed'] += 1
self.processing_stats['total_records'] += len(df)
else:
self.processing_stats['files_failed'] += 1
except Exception as e:
logger.error(f"Failed to process {file_path.name}: {e}")
self.processing_stats['files_failed'] += 1
if not dataframes:
logger.error("No dataframes were successfully loaded")
return False
# Merge all dataframes
logger.info("Starting merge process...")
self.merged_data = self.merge_dataframes_by_slug_datetime(dataframes)
if self.merged_data.empty:
logger.error("Merged dataframe is empty")
return False
# Update final stats
self.processing_stats['date_range'] = {
'start': str(self.merged_data.index.min()),
'end': str(self.merged_data.index.max()),
'total_days': (self.merged_data.index.max() - self.merged_data.index.min()).days
}
logger.info("All files processed successfully")
return True
except Exception as e:
logger.error(f"Failed to process files: {e}")
return False
def save_merged_features(self, filename: str = "merged_features.parquet") -> bool:
"""
Save the merged features to a parquet file with comprehensive null filling
Args:
filename: Output filename
Returns:
True if successful, False otherwise
"""
if self.merged_data is None or self.merged_data.empty:
logger.error("No merged data to save")
return False
try:
output_path = self.output_dir / filename
# Clean up the dataframe before saving
cleaned_df = self.merged_data.copy()
# Remove any completely null columns
null_columns = cleaned_df.columns[cleaned_df.isnull().all()].tolist()
if null_columns:
logger.info(f"Removing {len(null_columns)} completely null columns: {null_columns}")
cleaned_df = cleaned_df.dropna(axis=1, how='all')
# Apply comprehensive null filling strategy
logger.info("Applying comprehensive null filling...")
cleaned_df = self.fill_missing_values(cleaned_df)
# Remove unwanted columns
columns_to_remove = ['metric', 'source_file', 'category', 'development_alternative_slug_used']
existing_cols_to_remove = [col for col in columns_to_remove if col in cleaned_df.columns]
if existing_cols_to_remove:
logger.info(f"Removing unwanted columns: {existing_cols_to_remove}")
cleaned_df = cleaned_df.drop(columns=existing_cols_to_remove)
# Ensure all slugs are in uppercase format
logger.info("Ensuring all slugs are in uppercase format...")
cleaned_df['slug'] = cleaned_df['slug'].apply(lambda x: x.upper() if isinstance(x, str) else x)
# Fix data type issues for parquet compatibility
logger.info("Fixing data types for parquet compatibility...")
for col in cleaned_df.columns:
if cleaned_df[col].dtype == 'object':
# Check if column contains mixed types
sample_values = cleaned_df[col].dropna().head(100)
if len(sample_values) > 0:
# If it looks like it should be numeric, convert it
try:
pd.to_numeric(sample_values, errors='raise')
# If no error, convert the entire column
cleaned_df[col] = pd.to_numeric(cleaned_df[col], errors='coerce')
logger.debug(f"Converted {col} to numeric")
except (ValueError, TypeError):
# If conversion fails, ensure it's all strings
cleaned_df[col] = cleaned_df[col].astype(str)
logger.debug(f"Converted {col} to string")
# Sort by datetime and slug
cleaned_df = cleaned_df.sort_index()
cleaned_df = cleaned_df.sort_values(['slug'], kind='mergesort')
# Final data quality check
remaining_nulls = cleaned_df.isnull().sum().sum()
if remaining_nulls > 0:
logger.warning(f"Warning: {remaining_nulls} null values remain after filling")
# Log columns with remaining nulls
null_cols = cleaned_df.columns[cleaned_df.isnull().any()].tolist()
logger.warning(f"Columns with remaining nulls: {null_cols}")
else:
logger.info("✓ All null values successfully filled")
# Save to parquet with error handling
try:
cleaned_df.to_parquet(output_path, compression='snappy')
except Exception as parquet_error:
logger.error(f"Parquet save failed: {parquet_error}")
# Try to identify problematic columns
logger.info("Analyzing columns for parquet compatibility...")
for col in cleaned_df.columns:
try:
test_df = cleaned_df[[col]].copy()
test_df.to_parquet(output_path.with_suffix('.test.parquet'))
output_path.with_suffix('.test.parquet').unlink() # Clean up test file
except Exception as col_error:
logger.error(f"Column {col} causing issues: {col_error}")
# Force convert problematic column to string
cleaned_df[col] = cleaned_df[col].astype(str)
logger.info(f"Converted problematic column {col} to string")
# Try saving again
cleaned_df.to_parquet(output_path, compression='snappy')
logger.info(f"Merged features saved to {output_path}")
logger.info(f"Final dataset: {len(cleaned_df)} records, {len(cleaned_df.columns)} columns")
logger.info(f"Data completeness: {100 - (remaining_nulls / (len(cleaned_df) * len(cleaned_df.columns)) * 100):.2f}%")
return True
except Exception as e:
logger.error(f"Failed to save merged features: {e}")
return False
def generate_summary_report(self) -> Dict:
"""
Generate a comprehensive summary report
Returns:
Summary dictionary
"""
summary = {
'processing_timestamp': datetime.now().isoformat(),
'files_statistics': {
'files_found': self.processing_stats['files_found'],
'files_processed': self.processing_stats['files_processed'],
'files_failed': self.processing_stats['files_failed'],
'success_rate': f"{(self.processing_stats['files_processed'] / max(1, self.processing_stats['files_found'])) * 100:.1f}%"
},
'data_statistics': {
'total_records': self.processing_stats['total_records'],
'unique_slugs': list(self.processing_stats['unique_slugs']),
'categories_found': list(self.processing_stats['categories']),
'date_range': self.processing_stats['date_range']
}
}
if self.merged_data is not None:
summary['merged_statistics'] = {
'final_records': len(self.merged_data),
'final_columns': len(self.merged_data.columns),
'memory_usage_mb': f"{self.merged_data.memory_usage(deep=True).sum() / 1024 / 1024:.2f}",
'slug_distribution': self.merged_data['slug'].value_counts().to_dict(),
'null_percentage': f"{(self.merged_data.isnull().sum().sum() / (len(self.merged_data) * len(self.merged_data.columns))) * 100:.2f}%"
}
return summary
def print_summary(self):
"""Print a comprehensive summary of the merge process"""
summary = self.generate_summary_report()
print("\n" + "="*60)
print("SANTIMENT DATA MERGER SUMMARY")
print("="*60)
# File statistics
print(f"\nFile Processing:")
print(f" Files found: {summary['files_statistics']['files_found']}")
print(f" Files processed: {summary['files_statistics']['files_processed']}")
print(f" Files failed: {summary['files_statistics']['files_failed']}")
print(f" Success rate: {summary['files_statistics']['success_rate']}")
# Data statistics
print(f"\nData Overview:")
print(f" Total records processed: {summary['data_statistics']['total_records']:,}")
print(f" Unique assets (slugs): {len(summary['data_statistics']['unique_slugs'])}")
print(f" Categories found: {', '.join(summary['data_statistics']['categories_found'])}")
if summary['data_statistics']['date_range']:
print(f" Date range: {summary['data_statistics']['date_range']['start']} to {summary['data_statistics']['date_range']['end']}")
print(f" Total days: {summary['data_statistics']['date_range']['total_days']}")
# Merged statistics
if 'merged_statistics' in summary:
print(f"\nMerged Dataset:")
print(f" Final records: {summary['merged_statistics']['final_records']:,}")
print(f" Final columns: {summary['merged_statistics']['final_columns']}")
print(f" Memory usage: {summary['merged_statistics']['memory_usage_mb']} MB")
print(f" Data completeness: {100 - float(summary['merged_statistics']['null_percentage'].rstrip('%')):.1f}%")
# Show top assets by record count
print(f"\nTop Assets by Record Count:")
slug_dist = summary['merged_statistics']['slug_distribution']
for slug, count in list(slug_dist.items())[:5]:
print(f" {slug}: {count:,} records")
print("="*60)
def main():
"""Main function to run the Santiment data merger"""
logger.info("Starting Santiment Data Merger...")
# Initialize the merger
merger = SantimentDataMerger(
source_dir="data/santiment",
output_dir="data/santiment",
time_tolerance_hours=1
)
try:
# Process all files
success = merger.process_all_files()
if not success:
logger.error("Failed to process Santiment files")
return False
# If we only created a placeholder, treat as successful and skip saving/summary
if merger.placeholder_created:
logger.info("Placeholder Santiment dataset created; skipping save and summary.")
return True
# Save merged features
save_success = merger.save_merged_features("merged_features.parquet")
if not save_success:
logger.error("Failed to save merged features")
return False
# Print summary
merger.print_summary()
# Save summary report
summary = merger.generate_summary_report()
summary_path = Path("data/santiment") / "merge_summary.json"
import json
with open(summary_path, 'w') as f:
json.dump(summary, f, indent=2, default=str)
logger.info(f"Summary report saved to {summary_path}")
logger.info("Santiment data merge completed successfully!")
return True
except Exception as e:
logger.error(f"Santiment data merge failed: {e}")
return False
if __name__ == "__main__":
main()
|