WIP oclcv in storage. migrate do duckdb
This commit is contained in:
Dobromir Popov
@@ -67,6 +67,14 @@ except ImportError:
|
||||
UNIFIED_STORAGE_AVAILABLE = False
|
||||
logger.warning("Unified storage components not available")
|
||||
|
||||
# Import DuckDB storage
|
||||
try:
|
||||
from .duckdb_storage import DuckDBStorage
|
||||
DUCKDB_STORAGE_AVAILABLE = True
|
||||
except ImportError:
|
||||
DUCKDB_STORAGE_AVAILABLE = False
|
||||
logger.warning("DuckDB storage not available")
|
||||
|
||||
@dataclass
|
||||
class PivotBounds:
|
||||
"""Pivot-based normalization bounds derived from Williams Market Structure"""
|
||||
@@ -142,15 +150,10 @@ class DataProvider:
|
||||
def __init__(self, symbols: List[str] = None, timeframes: List[str] = None):
|
||||
"""Initialize the data provider"""
|
||||
self.config = get_config()
|
||||
# Fixed symbols and timeframes for caching
|
||||
# Fixed symbols and timeframes
|
||||
self.symbols = ['ETH/USDT', 'BTC/USDT']
|
||||
self.timeframes = ['1s', '1m', '1h', '1d']
|
||||
|
||||
# Cache settings (initialize first)
|
||||
self.cache_enabled = True
|
||||
self.cache_dir = Path('cache')
|
||||
self.cache_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# Data storage - cached OHLCV data (1500 candles each)
|
||||
self.cached_data = {} # {symbol: {timeframe: DataFrame}}
|
||||
self.real_time_data = {} # {symbol: {timeframe: deque}}
|
||||
@@ -176,11 +179,7 @@ class DataProvider:
|
||||
|
||||
# Pivot-based normalization system
|
||||
self.pivot_bounds: Dict[str, PivotBounds] = {} # {symbol: PivotBounds}
|
||||
self.pivot_cache_dir = self.cache_dir / 'pivot_bounds'
|
||||
self.pivot_cache_dir.mkdir(parents=True, exist_ok=True)
|
||||
self.pivot_refresh_interval = timedelta(days=1) # Refresh pivot bounds daily
|
||||
self.monthly_data_cache_dir = self.cache_dir / 'monthly_1s_data'
|
||||
self.monthly_data_cache_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# Enhanced WebSocket integration
|
||||
self.enhanced_cob_websocket: Optional[EnhancedCOBWebSocket] = None
|
||||
@@ -266,11 +265,16 @@ class DataProvider:
|
||||
self.unified_storage: Optional['UnifiedDataProviderExtension'] = None
|
||||
self._unified_storage_enabled = False
|
||||
|
||||
# Auto-fix corrupted cache files on startup
|
||||
self._auto_fix_corrupted_cache()
|
||||
# DuckDB storage - unified storage with native Parquet support
|
||||
self.duckdb_storage: Optional[DuckDBStorage] = None
|
||||
if DUCKDB_STORAGE_AVAILABLE:
|
||||
try:
|
||||
self.duckdb_storage = DuckDBStorage()
|
||||
logger.info("✅ DuckDB storage initialized (unified Parquet + SQL)")
|
||||
except Exception as e:
|
||||
logger.warning(f"Could not initialize DuckDB storage: {e}")
|
||||
|
||||
# Load existing pivot bounds from cache
|
||||
self._load_all_pivot_bounds()
|
||||
# Pivot bounds will be calculated on demand
|
||||
|
||||
# COB (Consolidated Order Book) data system using WebSocket
|
||||
self.cob_integration: Optional[COBIntegration] = None
|
||||
@@ -1488,11 +1492,18 @@ class DataProvider:
|
||||
logger.error(f"Error getting market state at time: {e}")
|
||||
return {}
|
||||
|
||||
def get_historical_data(self, symbol: str, timeframe: str, limit: int = 1000, refresh: bool = False) -> Optional[pd.DataFrame]:
|
||||
def get_historical_data(self, symbol: str, timeframe: str, limit: int = 1000, refresh: bool = False, allow_stale_cache: bool = False) -> Optional[pd.DataFrame]:
|
||||
"""Get historical OHLCV data.
|
||||
- Prefer cached data for low latency.
|
||||
- If cache is empty or refresh=True, fetch real data from exchanges.
|
||||
- Never generate synthetic data.
|
||||
|
||||
Args:
|
||||
symbol: Trading symbol
|
||||
timeframe: Timeframe
|
||||
limit: Number of candles to return
|
||||
refresh: Force refresh from exchange
|
||||
allow_stale_cache: Allow loading stale cache (for startup performance)
|
||||
"""
|
||||
try:
|
||||
# Serve from cache when available
|
||||
@@ -1501,6 +1512,17 @@ class DataProvider:
|
||||
if not cached_df.empty and not refresh:
|
||||
return cached_df.tail(limit)
|
||||
|
||||
# Try loading from DuckDB first (fast Parquet queries)
|
||||
if allow_stale_cache:
|
||||
cached_df = self._load_from_duckdb(symbol, timeframe, limit=1500)
|
||||
if cached_df is not None and not cached_df.empty:
|
||||
logger.info(f"Loaded {len(cached_df)} candles from DuckDB for {symbol} {timeframe} (startup mode)")
|
||||
# Store in memory cache
|
||||
if symbol not in self.cached_data:
|
||||
self.cached_data[symbol] = {}
|
||||
self.cached_data[symbol][timeframe] = cached_df.tail(1500)
|
||||
return cached_df.tail(limit)
|
||||
|
||||
# Cache empty or refresh requested: fetch real data now
|
||||
df = self._fetch_from_binance(symbol, timeframe, limit)
|
||||
if (df is None or df.empty):
|
||||
@@ -1508,7 +1530,15 @@ class DataProvider:
|
||||
|
||||
if df is not None and not df.empty:
|
||||
df = self._ensure_datetime_index(df)
|
||||
# Store/merge into cache
|
||||
|
||||
# Store in DuckDB (Parquet + SQL in one)
|
||||
if self.duckdb_storage:
|
||||
try:
|
||||
self.duckdb_storage.store_ohlcv_data(symbol, timeframe, df)
|
||||
except Exception as e:
|
||||
logger.warning(f"Could not store data in DuckDB: {e}")
|
||||
|
||||
# Store/merge into memory cache (keep last 1500 candles for fast access)
|
||||
if symbol not in self.cached_data:
|
||||
self.cached_data[symbol] = {}
|
||||
if timeframe not in self.cached_data[symbol] or self.cached_data[symbol][timeframe].empty:
|
||||
@@ -1518,7 +1548,8 @@ class DataProvider:
|
||||
combined_df = combined_df[~combined_df.index.duplicated(keep='last')]
|
||||
combined_df = combined_df.sort_index()
|
||||
self.cached_data[symbol][timeframe] = combined_df.tail(1500)
|
||||
logger.info(f"Cached {len(self.cached_data[symbol][timeframe])} candles for {symbol} {timeframe}")
|
||||
|
||||
logger.info(f"Stored {len(df)} candles for {symbol} {timeframe} (DuckDB + memory cache)")
|
||||
return self.cached_data[symbol][timeframe].tail(limit)
|
||||
|
||||
logger.warning(f"No real data available for {symbol} {timeframe} at request time")
|
||||
@@ -2973,71 +3004,33 @@ class DataProvider:
|
||||
logger.debug(f"Error calculating RSI: {e}")
|
||||
return 50.0 # Default neutral value
|
||||
|
||||
def _load_from_cache(self, symbol: str, timeframe: str) -> Optional[pd.DataFrame]:
|
||||
"""Load data from cache"""
|
||||
try:
|
||||
cache_file = self.cache_dir / f"{symbol.replace('/', '')}_{timeframe}.parquet"
|
||||
if cache_file.exists():
|
||||
# Check if cache is recent - stricter rules for startup
|
||||
cache_age = time.time() - cache_file.stat().st_mtime
|
||||
|
||||
# For 1m data, use cache only if less than 5 minutes old to avoid gaps
|
||||
if timeframe == '1m':
|
||||
max_age = 300 # 5 minutes
|
||||
else:
|
||||
max_age = 3600 # 1 hour for other timeframes
|
||||
|
||||
if cache_age < max_age:
|
||||
try:
|
||||
df = pd.read_parquet(cache_file)
|
||||
# Ensure cached data has proper timezone (UTC to match COB WebSocket data)
|
||||
if not df.empty and 'timestamp' in df.columns:
|
||||
if df['timestamp'].dt.tz is None:
|
||||
# If no timezone info, assume UTC and keep in UTC
|
||||
df['timestamp'] = pd.to_datetime(df['timestamp'], utc=True)
|
||||
elif str(df['timestamp'].dt.tz) != 'UTC':
|
||||
# Convert to UTC if different timezone
|
||||
df['timestamp'] = df['timestamp'].dt.tz_convert('UTC')
|
||||
logger.debug(f"Loaded {len(df)} rows from cache for {symbol} {timeframe} (age: {cache_age/60:.1f}min)")
|
||||
return df
|
||||
except Exception as parquet_e:
|
||||
# Handle corrupted Parquet file - expanded error detection
|
||||
error_str = str(parquet_e).lower()
|
||||
corrupted_indicators = [
|
||||
"parquet magic bytes not found",
|
||||
"corrupted",
|
||||
"couldn't deserialize thrift",
|
||||
"don't know what type",
|
||||
"invalid parquet file",
|
||||
"unexpected end of file",
|
||||
"invalid metadata"
|
||||
]
|
||||
|
||||
if any(indicator in error_str for indicator in corrupted_indicators):
|
||||
logger.warning(f"Corrupted Parquet cache file for {symbol} {timeframe}, removing and returning None: {parquet_e}")
|
||||
try:
|
||||
cache_file.unlink() # Delete corrupted file
|
||||
logger.info(f"Deleted corrupted cache file: {cache_file}")
|
||||
except Exception as delete_e:
|
||||
logger.error(f"Failed to delete corrupted cache file: {delete_e}")
|
||||
return None
|
||||
else:
|
||||
raise parquet_e
|
||||
else:
|
||||
logger.debug(f"Cache for {symbol} {timeframe} is too old ({cache_age/60:.1f}min > {max_age/60:.1f}min)")
|
||||
def _load_from_duckdb(self, symbol: str, timeframe: str, limit: int = 1500) -> Optional[pd.DataFrame]:
|
||||
"""Load data from DuckDB storage
|
||||
|
||||
Args:
|
||||
symbol: Trading symbol
|
||||
timeframe: Timeframe
|
||||
limit: Number of candles to load
|
||||
"""
|
||||
if not self.duckdb_storage:
|
||||
return None
|
||||
except Exception as e:
|
||||
logger.warning(f"Error loading cache for {symbol} {timeframe}: {e}")
|
||||
return None
|
||||
|
||||
def _save_to_cache(self, df: pd.DataFrame, symbol: str, timeframe: str):
|
||||
"""Save data to cache"""
|
||||
|
||||
try:
|
||||
cache_file = self.cache_dir / f"{symbol.replace('/', '')}_{timeframe}.parquet"
|
||||
df.to_parquet(cache_file, index=False)
|
||||
logger.debug(f"Saved {len(df)} rows to cache for {symbol} {timeframe}")
|
||||
df = self.duckdb_storage.get_ohlcv_data(
|
||||
symbol=symbol,
|
||||
timeframe=timeframe,
|
||||
limit=limit
|
||||
)
|
||||
|
||||
if df is not None and not df.empty:
|
||||
logger.debug(f"Loaded {len(df)} candles from DuckDB for {symbol} {timeframe}")
|
||||
return df
|
||||
|
||||
return None
|
||||
|
||||
except Exception as e:
|
||||
logger.warning(f"Error saving cache for {symbol} {timeframe}: {e}")
|
||||
logger.warning(f"Error loading from DuckDB for {symbol} {timeframe}: {e}")
|
||||
return None
|
||||
|
||||
async def start_real_time_streaming(self):
|
||||
"""Start real-time data streaming using COBIntegration"""
|
||||
|
||||
429
core/duckdb_storage.py
Normal file
429
core/duckdb_storage.py
Normal file
@@ -0,0 +1,429 @@
|
||||
"""
|
||||
DuckDB Storage - Unified Storage with Native Parquet Support
|
||||
|
||||
DuckDB provides the best of both worlds:
|
||||
- Native Parquet support (query files directly)
|
||||
- Full SQL capabilities (complex queries)
|
||||
- Columnar storage (fast analytics)
|
||||
- Zero-copy reads (extremely fast)
|
||||
- Embedded database (no server)
|
||||
|
||||
This replaces the dual SQLite + Parquet system with a single unified solution.
|
||||
"""
|
||||
|
||||
import duckdb
|
||||
import logging
|
||||
import pandas as pd
|
||||
from datetime import datetime
|
||||
from pathlib import Path
|
||||
from typing import Dict, List, Optional, Tuple, Any
|
||||
import json
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class DuckDBStorage:
|
||||
"""Unified storage using DuckDB with native Parquet support"""
|
||||
|
||||
def __init__(self, db_path: str = "cache/trading_data.duckdb"):
|
||||
"""Initialize DuckDB storage"""
|
||||
self.db_path = Path(db_path)
|
||||
self.db_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# Parquet storage directory
|
||||
self.parquet_dir = self.db_path.parent / "parquet_store"
|
||||
self.parquet_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# Connect to DuckDB
|
||||
self.conn = duckdb.connect(str(self.db_path))
|
||||
|
||||
# Initialize schema
|
||||
self._init_schema()
|
||||
|
||||
logger.info(f"DuckDB storage initialized: {self.db_path}")
|
||||
logger.info(f"Parquet storage: {self.parquet_dir}")
|
||||
|
||||
def _init_schema(self):
|
||||
"""Initialize database schema with Parquet integration"""
|
||||
|
||||
# Create annotations table (metadata only)
|
||||
self.conn.execute("""
|
||||
CREATE TABLE IF NOT EXISTS annotations (
|
||||
annotation_id VARCHAR PRIMARY KEY,
|
||||
symbol VARCHAR NOT NULL,
|
||||
timeframe VARCHAR NOT NULL,
|
||||
direction VARCHAR NOT NULL,
|
||||
entry_timestamp BIGINT NOT NULL,
|
||||
entry_price DOUBLE NOT NULL,
|
||||
exit_timestamp BIGINT NOT NULL,
|
||||
exit_price DOUBLE NOT NULL,
|
||||
profit_loss_pct DOUBLE NOT NULL,
|
||||
notes TEXT,
|
||||
created_at BIGINT NOT NULL,
|
||||
market_context JSON,
|
||||
model_features JSON,
|
||||
pivot_data JSON,
|
||||
parquet_path VARCHAR
|
||||
)
|
||||
""")
|
||||
|
||||
# Create cache metadata table
|
||||
self.conn.execute("""
|
||||
CREATE TABLE IF NOT EXISTS cache_metadata (
|
||||
symbol VARCHAR NOT NULL,
|
||||
timeframe VARCHAR NOT NULL,
|
||||
parquet_path VARCHAR NOT NULL,
|
||||
first_timestamp BIGINT NOT NULL,
|
||||
last_timestamp BIGINT NOT NULL,
|
||||
candle_count INTEGER NOT NULL,
|
||||
last_update BIGINT NOT NULL,
|
||||
PRIMARY KEY (symbol, timeframe)
|
||||
)
|
||||
""")
|
||||
|
||||
logger.info("DuckDB schema initialized")
|
||||
|
||||
def store_ohlcv_data(self, symbol: str, timeframe: str, df: pd.DataFrame) -> int:
|
||||
"""
|
||||
Store OHLCV data as Parquet file and register in DuckDB
|
||||
|
||||
Args:
|
||||
symbol: Trading symbol
|
||||
timeframe: Timeframe
|
||||
df: DataFrame with OHLCV data
|
||||
|
||||
Returns:
|
||||
Number of rows stored
|
||||
"""
|
||||
if df is None or df.empty:
|
||||
return 0
|
||||
|
||||
try:
|
||||
# Prepare data
|
||||
df_copy = df.copy()
|
||||
|
||||
# Ensure timestamp column
|
||||
if 'timestamp' not in df_copy.columns:
|
||||
df_copy['timestamp'] = df_copy.index
|
||||
|
||||
# Convert timestamp to Unix milliseconds
|
||||
if pd.api.types.is_datetime64_any_dtype(df_copy['timestamp']):
|
||||
df_copy['timestamp'] = df_copy['timestamp'].astype('int64') // 10**6
|
||||
|
||||
# Add metadata
|
||||
df_copy['symbol'] = symbol
|
||||
df_copy['timeframe'] = timeframe
|
||||
|
||||
# Define parquet file path
|
||||
parquet_file = self.parquet_dir / f"{symbol.replace('/', '_')}_{timeframe}.parquet"
|
||||
|
||||
# Load existing data if file exists
|
||||
if parquet_file.exists():
|
||||
try:
|
||||
existing_df = pd.read_parquet(parquet_file)
|
||||
# Combine with new data
|
||||
df_copy = pd.concat([existing_df, df_copy], ignore_index=True)
|
||||
# Remove duplicates
|
||||
df_copy = df_copy.drop_duplicates(subset=['timestamp'], keep='last')
|
||||
df_copy = df_copy.sort_values('timestamp')
|
||||
except Exception as e:
|
||||
logger.warning(f"Could not load existing parquet: {e}")
|
||||
|
||||
# Save to parquet
|
||||
df_copy.to_parquet(parquet_file, index=False, compression='snappy')
|
||||
|
||||
# Update metadata in DuckDB
|
||||
first_ts = int(df_copy['timestamp'].min())
|
||||
last_ts = int(df_copy['timestamp'].max())
|
||||
count = len(df_copy)
|
||||
now_ts = int(datetime.now().timestamp() * 1000)
|
||||
|
||||
self.conn.execute("""
|
||||
INSERT OR REPLACE INTO cache_metadata
|
||||
(symbol, timeframe, parquet_path, first_timestamp, last_timestamp, candle_count, last_update)
|
||||
VALUES (?, ?, ?, ?, ?, ?, ?)
|
||||
""", (symbol, timeframe, str(parquet_file), first_ts, last_ts, count, now_ts))
|
||||
|
||||
logger.info(f"Stored {len(df)} candles for {symbol} {timeframe} in Parquet (total: {count})")
|
||||
return len(df)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error storing OHLCV data: {e}")
|
||||
import traceback
|
||||
traceback.print_exc()
|
||||
return 0
|
||||
|
||||
def get_ohlcv_data(self, symbol: str, timeframe: str,
|
||||
start_time: Optional[datetime] = None,
|
||||
end_time: Optional[datetime] = None,
|
||||
limit: Optional[int] = None) -> Optional[pd.DataFrame]:
|
||||
"""
|
||||
Query OHLCV data directly from Parquet using DuckDB
|
||||
|
||||
Args:
|
||||
symbol: Trading symbol
|
||||
timeframe: Timeframe
|
||||
start_time: Start time filter
|
||||
end_time: End time filter
|
||||
limit: Maximum number of candles
|
||||
|
||||
Returns:
|
||||
DataFrame with OHLCV data
|
||||
"""
|
||||
try:
|
||||
# Get parquet file path from metadata
|
||||
result = self.conn.execute("""
|
||||
SELECT parquet_path FROM cache_metadata
|
||||
WHERE symbol = ? AND timeframe = ?
|
||||
""", (symbol, timeframe)).fetchone()
|
||||
|
||||
if not result:
|
||||
logger.debug(f"No data found for {symbol} {timeframe}")
|
||||
return None
|
||||
|
||||
parquet_path = result[0]
|
||||
|
||||
if not Path(parquet_path).exists():
|
||||
logger.warning(f"Parquet file not found: {parquet_path}")
|
||||
return None
|
||||
|
||||
# Build query - DuckDB can query Parquet directly!
|
||||
query = f"""
|
||||
SELECT timestamp, open, high, low, close, volume
|
||||
FROM read_parquet('{parquet_path}')
|
||||
WHERE symbol = ? AND timeframe = ?
|
||||
"""
|
||||
params = [symbol, timeframe]
|
||||
|
||||
if start_time:
|
||||
query += " AND timestamp >= ?"
|
||||
params.append(int(start_time.timestamp() * 1000))
|
||||
|
||||
if end_time:
|
||||
query += " AND timestamp <= ?"
|
||||
params.append(int(end_time.timestamp() * 1000))
|
||||
|
||||
query += " ORDER BY timestamp DESC"
|
||||
|
||||
if limit:
|
||||
query += f" LIMIT {limit}"
|
||||
|
||||
# Execute query
|
||||
df = self.conn.execute(query, params).df()
|
||||
|
||||
if df.empty:
|
||||
return None
|
||||
|
||||
# Convert timestamp to datetime
|
||||
df['timestamp'] = pd.to_datetime(df['timestamp'], unit='ms', utc=True)
|
||||
df = df.set_index('timestamp')
|
||||
df = df.sort_index()
|
||||
|
||||
logger.debug(f"Retrieved {len(df)} candles for {symbol} {timeframe} from Parquet")
|
||||
return df
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error retrieving OHLCV data: {e}")
|
||||
import traceback
|
||||
traceback.print_exc()
|
||||
return None
|
||||
|
||||
def store_annotation(self, annotation_id: str, annotation_data: Dict[str, Any],
|
||||
market_snapshots: Dict[str, pd.DataFrame],
|
||||
model_predictions: Optional[List[Dict]] = None) -> bool:
|
||||
"""
|
||||
Store annotation with market snapshots as Parquet
|
||||
|
||||
Args:
|
||||
annotation_id: Unique annotation ID
|
||||
annotation_data: Annotation metadata
|
||||
market_snapshots: Dict of {timeframe: DataFrame} with OHLCV data
|
||||
model_predictions: List of model predictions
|
||||
|
||||
Returns:
|
||||
True if successful
|
||||
"""
|
||||
try:
|
||||
# Parse timestamps
|
||||
entry_time = annotation_data.get('entry', {}).get('timestamp')
|
||||
exit_time = annotation_data.get('exit', {}).get('timestamp')
|
||||
|
||||
if isinstance(entry_time, str):
|
||||
entry_time = datetime.fromisoformat(entry_time.replace('Z', '+00:00'))
|
||||
if isinstance(exit_time, str):
|
||||
exit_time = datetime.fromisoformat(exit_time.replace('Z', '+00:00'))
|
||||
|
||||
# Store market snapshots as Parquet
|
||||
annotation_parquet_dir = self.parquet_dir / "annotations" / annotation_id
|
||||
annotation_parquet_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
for timeframe, df in market_snapshots.items():
|
||||
if df is None or df.empty:
|
||||
continue
|
||||
|
||||
df_copy = df.copy()
|
||||
|
||||
# Ensure timestamp column
|
||||
if 'timestamp' not in df_copy.columns:
|
||||
df_copy['timestamp'] = df_copy.index
|
||||
|
||||
# Convert timestamp
|
||||
if pd.api.types.is_datetime64_any_dtype(df_copy['timestamp']):
|
||||
df_copy['timestamp'] = df_copy['timestamp'].astype('int64') // 10**6
|
||||
|
||||
# Save to parquet
|
||||
parquet_file = annotation_parquet_dir / f"{timeframe}.parquet"
|
||||
df_copy.to_parquet(parquet_file, index=False, compression='snappy')
|
||||
|
||||
# Store annotation metadata in DuckDB
|
||||
self.conn.execute("""
|
||||
INSERT OR REPLACE INTO annotations
|
||||
(annotation_id, symbol, timeframe, direction,
|
||||
entry_timestamp, entry_price, exit_timestamp, exit_price,
|
||||
profit_loss_pct, notes, created_at, market_context,
|
||||
model_features, pivot_data, parquet_path)
|
||||
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
|
||||
""", (
|
||||
annotation_id,
|
||||
annotation_data.get('symbol'),
|
||||
annotation_data.get('timeframe'),
|
||||
annotation_data.get('direction'),
|
||||
int(entry_time.timestamp() * 1000),
|
||||
annotation_data.get('entry', {}).get('price'),
|
||||
int(exit_time.timestamp() * 1000),
|
||||
annotation_data.get('exit', {}).get('price'),
|
||||
annotation_data.get('profit_loss_pct'),
|
||||
annotation_data.get('notes', ''),
|
||||
int(datetime.now().timestamp() * 1000),
|
||||
json.dumps(annotation_data.get('entry_market_state', {})),
|
||||
json.dumps(annotation_data.get('model_features', {})),
|
||||
json.dumps(annotation_data.get('pivot_data', {})),
|
||||
str(annotation_parquet_dir)
|
||||
))
|
||||
|
||||
logger.info(f"Stored annotation {annotation_id} with {len(market_snapshots)} timeframes")
|
||||
return True
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error storing annotation: {e}")
|
||||
import traceback
|
||||
traceback.print_exc()
|
||||
return False
|
||||
|
||||
def get_annotation(self, annotation_id: str) -> Optional[Dict[str, Any]]:
|
||||
"""
|
||||
Retrieve annotation with market snapshots from Parquet
|
||||
|
||||
Args:
|
||||
annotation_id: Annotation ID
|
||||
|
||||
Returns:
|
||||
Dict with annotation data and OHLCV snapshots
|
||||
"""
|
||||
try:
|
||||
# Get annotation metadata
|
||||
result = self.conn.execute("""
|
||||
SELECT * FROM annotations WHERE annotation_id = ?
|
||||
""", (annotation_id,)).fetchone()
|
||||
|
||||
if not result:
|
||||
return None
|
||||
|
||||
# Parse annotation data
|
||||
columns = [desc[0] for desc in self.conn.description]
|
||||
annotation = dict(zip(columns, result))
|
||||
|
||||
# Parse JSON fields
|
||||
annotation['market_context'] = json.loads(annotation.get('market_context', '{}'))
|
||||
annotation['model_features'] = json.loads(annotation.get('model_features', '{}'))
|
||||
annotation['pivot_data'] = json.loads(annotation.get('pivot_data', '{}'))
|
||||
|
||||
# Load OHLCV snapshots from Parquet
|
||||
parquet_dir = Path(annotation['parquet_path'])
|
||||
annotation['ohlcv_snapshots'] = {}
|
||||
|
||||
if parquet_dir.exists():
|
||||
for parquet_file in parquet_dir.glob('*.parquet'):
|
||||
timeframe = parquet_file.stem
|
||||
|
||||
# Query parquet directly with DuckDB
|
||||
df = self.conn.execute(f"""
|
||||
SELECT timestamp, open, high, low, close, volume
|
||||
FROM read_parquet('{parquet_file}')
|
||||
ORDER BY timestamp
|
||||
""").df()
|
||||
|
||||
if not df.empty:
|
||||
df['timestamp'] = pd.to_datetime(df['timestamp'], unit='ms', utc=True)
|
||||
df = df.set_index('timestamp')
|
||||
annotation['ohlcv_snapshots'][timeframe] = df
|
||||
|
||||
logger.info(f"Retrieved annotation {annotation_id} with {len(annotation['ohlcv_snapshots'])} timeframes")
|
||||
return annotation
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error retrieving annotation: {e}")
|
||||
return None
|
||||
|
||||
def query_sql(self, query: str, params: Optional[List] = None) -> pd.DataFrame:
|
||||
"""
|
||||
Execute arbitrary SQL query (including Parquet queries)
|
||||
|
||||
Args:
|
||||
query: SQL query
|
||||
params: Query parameters
|
||||
|
||||
Returns:
|
||||
DataFrame with results
|
||||
"""
|
||||
try:
|
||||
if params:
|
||||
result = self.conn.execute(query, params)
|
||||
else:
|
||||
result = self.conn.execute(query)
|
||||
|
||||
return result.df()
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error executing query: {e}")
|
||||
return pd.DataFrame()
|
||||
|
||||
def get_cache_stats(self) -> Dict[str, Any]:
|
||||
"""Get cache statistics"""
|
||||
try:
|
||||
# Get OHLCV stats
|
||||
ohlcv_stats = self.conn.execute("""
|
||||
SELECT symbol, timeframe, candle_count, first_timestamp, last_timestamp
|
||||
FROM cache_metadata
|
||||
ORDER BY symbol, timeframe
|
||||
""").df()
|
||||
|
||||
if not ohlcv_stats.empty:
|
||||
ohlcv_stats['first_timestamp'] = pd.to_datetime(ohlcv_stats['first_timestamp'], unit='ms')
|
||||
ohlcv_stats['last_timestamp'] = pd.to_datetime(ohlcv_stats['last_timestamp'], unit='ms')
|
||||
|
||||
# Get annotation count
|
||||
annotation_count = self.conn.execute("""
|
||||
SELECT COUNT(*) as count FROM annotations
|
||||
""").fetchone()[0]
|
||||
|
||||
# Get total candles
|
||||
total_candles = self.conn.execute("""
|
||||
SELECT SUM(candle_count) as total FROM cache_metadata
|
||||
""").fetchone()[0] or 0
|
||||
|
||||
return {
|
||||
'ohlcv_stats': ohlcv_stats.to_dict('records') if not ohlcv_stats.empty else [],
|
||||
'annotation_count': annotation_count,
|
||||
'total_candles': total_candles
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error getting cache stats: {e}")
|
||||
return {}
|
||||
|
||||
def close(self):
|
||||
"""Close database connection"""
|
||||
if self.conn:
|
||||
self.conn.close()
|
||||
logger.info("DuckDB connection closed")
|
||||
526
core/sqlite_storage.py
Normal file
526
core/sqlite_storage.py
Normal file
@@ -0,0 +1,526 @@
|
||||
"""
|
||||
SQLite Storage for Long-Term OHLCV Data and Annotation Replay
|
||||
|
||||
This module provides persistent storage for:
|
||||
1. OHLCV data for all timeframes (unlimited history)
|
||||
2. Complete annotation data with market context
|
||||
3. Model predictions and features at annotation time
|
||||
4. Efficient querying for historical replay
|
||||
|
||||
Parquet files are used for recent data (1500 candles) for speed.
|
||||
SQLite is used for long-term storage and annotation replay.
|
||||
"""
|
||||
|
||||
import sqlite3
|
||||
import logging
|
||||
import pandas as pd
|
||||
from datetime import datetime
|
||||
from pathlib import Path
|
||||
from typing import Dict, List, Optional, Tuple, Any
|
||||
import json
|
||||
import pickle
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class SQLiteStorage:
|
||||
"""SQLite storage for OHLCV data and annotations"""
|
||||
|
||||
def __init__(self, db_path: str = "cache/trading_data.db"):
|
||||
"""Initialize SQLite storage"""
|
||||
self.db_path = Path(db_path)
|
||||
self.db_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# Initialize database schema
|
||||
self._init_schema()
|
||||
|
||||
logger.info(f"SQLite storage initialized: {self.db_path}")
|
||||
|
||||
def _init_schema(self):
|
||||
"""Initialize database schema"""
|
||||
conn = sqlite3.connect(self.db_path)
|
||||
cursor = conn.cursor()
|
||||
|
||||
# OHLCV data table - stores all historical candles
|
||||
cursor.execute("""
|
||||
CREATE TABLE IF NOT EXISTS ohlcv_data (
|
||||
id INTEGER PRIMARY KEY AUTOINCREMENT,
|
||||
symbol TEXT NOT NULL,
|
||||
timeframe TEXT NOT NULL,
|
||||
timestamp INTEGER NOT NULL,
|
||||
open REAL NOT NULL,
|
||||
high REAL NOT NULL,
|
||||
low REAL NOT NULL,
|
||||
close REAL NOT NULL,
|
||||
volume REAL NOT NULL,
|
||||
created_at INTEGER NOT NULL,
|
||||
UNIQUE(symbol, timeframe, timestamp)
|
||||
)
|
||||
""")
|
||||
|
||||
# Indexes for fast queries
|
||||
cursor.execute("""
|
||||
CREATE INDEX IF NOT EXISTS idx_ohlcv_symbol_timeframe
|
||||
ON ohlcv_data(symbol, timeframe)
|
||||
""")
|
||||
cursor.execute("""
|
||||
CREATE INDEX IF NOT EXISTS idx_ohlcv_timestamp
|
||||
ON ohlcv_data(timestamp)
|
||||
""")
|
||||
cursor.execute("""
|
||||
CREATE INDEX IF NOT EXISTS idx_ohlcv_lookup
|
||||
ON ohlcv_data(symbol, timeframe, timestamp)
|
||||
""")
|
||||
|
||||
# Annotations table - stores complete annotation data
|
||||
cursor.execute("""
|
||||
CREATE TABLE IF NOT EXISTS annotations (
|
||||
annotation_id TEXT PRIMARY KEY,
|
||||
symbol TEXT NOT NULL,
|
||||
timeframe TEXT NOT NULL,
|
||||
direction TEXT NOT NULL,
|
||||
entry_timestamp INTEGER NOT NULL,
|
||||
entry_price REAL NOT NULL,
|
||||
exit_timestamp INTEGER NOT NULL,
|
||||
exit_price REAL NOT NULL,
|
||||
profit_loss_pct REAL NOT NULL,
|
||||
notes TEXT,
|
||||
created_at INTEGER NOT NULL,
|
||||
market_context TEXT,
|
||||
model_features TEXT,
|
||||
pivot_data TEXT
|
||||
)
|
||||
""")
|
||||
|
||||
# Annotation OHLCV snapshots - stores market data at annotation time
|
||||
cursor.execute("""
|
||||
CREATE TABLE IF NOT EXISTS annotation_ohlcv_snapshots (
|
||||
id INTEGER PRIMARY KEY AUTOINCREMENT,
|
||||
annotation_id TEXT NOT NULL,
|
||||
timeframe TEXT NOT NULL,
|
||||
timestamp INTEGER NOT NULL,
|
||||
open REAL NOT NULL,
|
||||
high REAL NOT NULL,
|
||||
low REAL NOT NULL,
|
||||
close REAL NOT NULL,
|
||||
volume REAL NOT NULL,
|
||||
FOREIGN KEY (annotation_id) REFERENCES annotations(annotation_id),
|
||||
UNIQUE(annotation_id, timeframe, timestamp)
|
||||
)
|
||||
""")
|
||||
|
||||
cursor.execute("""
|
||||
CREATE INDEX IF NOT EXISTS idx_annotation_snapshots
|
||||
ON annotation_ohlcv_snapshots(annotation_id, timeframe)
|
||||
""")
|
||||
|
||||
# Model predictions table - stores model outputs at annotation time
|
||||
cursor.execute("""
|
||||
CREATE TABLE IF NOT EXISTS annotation_model_predictions (
|
||||
id INTEGER PRIMARY KEY AUTOINCREMENT,
|
||||
annotation_id TEXT NOT NULL,
|
||||
model_name TEXT NOT NULL,
|
||||
timestamp INTEGER NOT NULL,
|
||||
prediction TEXT NOT NULL,
|
||||
confidence REAL,
|
||||
features TEXT,
|
||||
FOREIGN KEY (annotation_id) REFERENCES annotations(annotation_id)
|
||||
)
|
||||
""")
|
||||
|
||||
# Cache metadata table - tracks what data we have
|
||||
cursor.execute("""
|
||||
CREATE TABLE IF NOT EXISTS cache_metadata (
|
||||
symbol TEXT NOT NULL,
|
||||
timeframe TEXT NOT NULL,
|
||||
first_timestamp INTEGER NOT NULL,
|
||||
last_timestamp INTEGER NOT NULL,
|
||||
candle_count INTEGER NOT NULL,
|
||||
last_update INTEGER NOT NULL,
|
||||
PRIMARY KEY (symbol, timeframe)
|
||||
)
|
||||
""")
|
||||
|
||||
conn.commit()
|
||||
conn.close()
|
||||
|
||||
logger.info("SQLite schema initialized")
|
||||
|
||||
def store_ohlcv_data(self, symbol: str, timeframe: str, df: pd.DataFrame) -> int:
|
||||
"""
|
||||
Store OHLCV data in SQLite
|
||||
|
||||
Args:
|
||||
symbol: Trading symbol
|
||||
timeframe: Timeframe
|
||||
df: DataFrame with OHLCV data (timestamp as index or column)
|
||||
|
||||
Returns:
|
||||
Number of rows inserted
|
||||
"""
|
||||
if df is None or df.empty:
|
||||
return 0
|
||||
|
||||
try:
|
||||
conn = sqlite3.connect(self.db_path)
|
||||
|
||||
# Prepare data
|
||||
df_copy = df.copy()
|
||||
|
||||
# Ensure timestamp column exists
|
||||
if 'timestamp' not in df_copy.columns:
|
||||
df_copy['timestamp'] = df_copy.index
|
||||
|
||||
# Convert timestamp to Unix milliseconds
|
||||
if pd.api.types.is_datetime64_any_dtype(df_copy['timestamp']):
|
||||
df_copy['timestamp'] = df_copy['timestamp'].astype('int64') // 10**6
|
||||
|
||||
# Add metadata
|
||||
df_copy['symbol'] = symbol
|
||||
df_copy['timeframe'] = timeframe
|
||||
df_copy['created_at'] = int(datetime.now().timestamp() * 1000)
|
||||
|
||||
# Select columns in correct order
|
||||
columns = ['symbol', 'timeframe', 'timestamp', 'open', 'high', 'low', 'close', 'volume', 'created_at']
|
||||
df_insert = df_copy[columns]
|
||||
|
||||
# Insert data (ignore duplicates)
|
||||
df_insert.to_sql('ohlcv_data', conn, if_exists='append', index=False)
|
||||
|
||||
# Update metadata
|
||||
cursor = conn.cursor()
|
||||
cursor.execute("""
|
||||
INSERT OR REPLACE INTO cache_metadata
|
||||
(symbol, timeframe, first_timestamp, last_timestamp, candle_count, last_update)
|
||||
VALUES (?, ?,
|
||||
COALESCE((SELECT MIN(timestamp) FROM ohlcv_data WHERE symbol=? AND timeframe=?), ?),
|
||||
COALESCE((SELECT MAX(timestamp) FROM ohlcv_data WHERE symbol=? AND timeframe=?), ?),
|
||||
(SELECT COUNT(*) FROM ohlcv_data WHERE symbol=? AND timeframe=?),
|
||||
?)
|
||||
""", (
|
||||
symbol, timeframe,
|
||||
symbol, timeframe, df_copy['timestamp'].min(),
|
||||
symbol, timeframe, df_copy['timestamp'].max(),
|
||||
symbol, timeframe,
|
||||
int(datetime.now().timestamp() * 1000)
|
||||
))
|
||||
|
||||
conn.commit()
|
||||
rows_inserted = len(df_insert)
|
||||
|
||||
conn.close()
|
||||
|
||||
logger.info(f"Stored {rows_inserted} candles for {symbol} {timeframe} in SQLite")
|
||||
return rows_inserted
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error storing OHLCV data in SQLite: {e}")
|
||||
return 0
|
||||
|
||||
def get_ohlcv_data(self, symbol: str, timeframe: str,
|
||||
start_time: Optional[datetime] = None,
|
||||
end_time: Optional[datetime] = None,
|
||||
limit: Optional[int] = None) -> Optional[pd.DataFrame]:
|
||||
"""
|
||||
Retrieve OHLCV data from SQLite
|
||||
|
||||
Args:
|
||||
symbol: Trading symbol
|
||||
timeframe: Timeframe
|
||||
start_time: Start time filter
|
||||
end_time: End time filter
|
||||
limit: Maximum number of candles
|
||||
|
||||
Returns:
|
||||
DataFrame with OHLCV data
|
||||
"""
|
||||
try:
|
||||
conn = sqlite3.connect(self.db_path)
|
||||
|
||||
# Build query
|
||||
query = """
|
||||
SELECT timestamp, open, high, low, close, volume
|
||||
FROM ohlcv_data
|
||||
WHERE symbol = ? AND timeframe = ?
|
||||
"""
|
||||
params = [symbol, timeframe]
|
||||
|
||||
if start_time:
|
||||
query += " AND timestamp >= ?"
|
||||
params.append(int(start_time.timestamp() * 1000))
|
||||
|
||||
if end_time:
|
||||
query += " AND timestamp <= ?"
|
||||
params.append(int(end_time.timestamp() * 1000))
|
||||
|
||||
query += " ORDER BY timestamp DESC"
|
||||
|
||||
if limit:
|
||||
query += f" LIMIT {limit}"
|
||||
|
||||
# Execute query
|
||||
df = pd.read_sql_query(query, conn, params=params)
|
||||
|
||||
conn.close()
|
||||
|
||||
if df.empty:
|
||||
return None
|
||||
|
||||
# Convert timestamp to datetime
|
||||
df['timestamp'] = pd.to_datetime(df['timestamp'], unit='ms', utc=True)
|
||||
df = df.set_index('timestamp')
|
||||
df = df.sort_index()
|
||||
|
||||
logger.debug(f"Retrieved {len(df)} candles for {symbol} {timeframe} from SQLite")
|
||||
return df
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error retrieving OHLCV data from SQLite: {e}")
|
||||
return None
|
||||
|
||||
def store_annotation(self, annotation_id: str, annotation_data: Dict[str, Any],
|
||||
market_snapshots: Dict[str, pd.DataFrame],
|
||||
model_predictions: Optional[List[Dict]] = None) -> bool:
|
||||
"""
|
||||
Store complete annotation with market context and model data
|
||||
|
||||
Args:
|
||||
annotation_id: Unique annotation ID
|
||||
annotation_data: Annotation metadata (entry, exit, symbol, etc.)
|
||||
market_snapshots: Dict of {timeframe: DataFrame} with OHLCV data
|
||||
model_predictions: List of model predictions at annotation time
|
||||
|
||||
Returns:
|
||||
True if successful
|
||||
"""
|
||||
try:
|
||||
conn = sqlite3.connect(self.db_path)
|
||||
cursor = conn.cursor()
|
||||
|
||||
# Parse timestamps
|
||||
entry_time = annotation_data.get('entry', {}).get('timestamp')
|
||||
exit_time = annotation_data.get('exit', {}).get('timestamp')
|
||||
|
||||
if isinstance(entry_time, str):
|
||||
entry_time = datetime.fromisoformat(entry_time.replace('Z', '+00:00'))
|
||||
if isinstance(exit_time, str):
|
||||
exit_time = datetime.fromisoformat(exit_time.replace('Z', '+00:00'))
|
||||
|
||||
# Store annotation metadata
|
||||
cursor.execute("""
|
||||
INSERT OR REPLACE INTO annotations
|
||||
(annotation_id, symbol, timeframe, direction,
|
||||
entry_timestamp, entry_price, exit_timestamp, exit_price,
|
||||
profit_loss_pct, notes, created_at, market_context, model_features, pivot_data)
|
||||
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
|
||||
""", (
|
||||
annotation_id,
|
||||
annotation_data.get('symbol'),
|
||||
annotation_data.get('timeframe'),
|
||||
annotation_data.get('direction'),
|
||||
int(entry_time.timestamp() * 1000),
|
||||
annotation_data.get('entry', {}).get('price'),
|
||||
int(exit_time.timestamp() * 1000),
|
||||
annotation_data.get('exit', {}).get('price'),
|
||||
annotation_data.get('profit_loss_pct'),
|
||||
annotation_data.get('notes', ''),
|
||||
int(datetime.now().timestamp() * 1000),
|
||||
json.dumps(annotation_data.get('entry_market_state', {})),
|
||||
json.dumps(annotation_data.get('model_features', {})),
|
||||
json.dumps(annotation_data.get('pivot_data', {}))
|
||||
))
|
||||
|
||||
# Store OHLCV snapshots for each timeframe
|
||||
for timeframe, df in market_snapshots.items():
|
||||
if df is None or df.empty:
|
||||
continue
|
||||
|
||||
df_copy = df.copy()
|
||||
|
||||
# Ensure timestamp column
|
||||
if 'timestamp' not in df_copy.columns:
|
||||
df_copy['timestamp'] = df_copy.index
|
||||
|
||||
# Convert timestamp to Unix milliseconds
|
||||
if pd.api.types.is_datetime64_any_dtype(df_copy['timestamp']):
|
||||
df_copy['timestamp'] = df_copy['timestamp'].astype('int64') // 10**6
|
||||
|
||||
# Insert each candle
|
||||
for _, row in df_copy.iterrows():
|
||||
cursor.execute("""
|
||||
INSERT OR REPLACE INTO annotation_ohlcv_snapshots
|
||||
(annotation_id, timeframe, timestamp, open, high, low, close, volume)
|
||||
VALUES (?, ?, ?, ?, ?, ?, ?, ?)
|
||||
""", (
|
||||
annotation_id,
|
||||
timeframe,
|
||||
int(row['timestamp']),
|
||||
float(row['open']),
|
||||
float(row['high']),
|
||||
float(row['low']),
|
||||
float(row['close']),
|
||||
float(row['volume'])
|
||||
))
|
||||
|
||||
# Store model predictions
|
||||
if model_predictions:
|
||||
for pred in model_predictions:
|
||||
cursor.execute("""
|
||||
INSERT INTO annotation_model_predictions
|
||||
(annotation_id, model_name, timestamp, prediction, confidence, features)
|
||||
VALUES (?, ?, ?, ?, ?, ?)
|
||||
""", (
|
||||
annotation_id,
|
||||
pred.get('model_name'),
|
||||
int(pred.get('timestamp', datetime.now().timestamp() * 1000)),
|
||||
json.dumps(pred.get('prediction')),
|
||||
pred.get('confidence'),
|
||||
json.dumps(pred.get('features', {}))
|
||||
))
|
||||
|
||||
conn.commit()
|
||||
conn.close()
|
||||
|
||||
logger.info(f"Stored annotation {annotation_id} with {len(market_snapshots)} timeframes in SQLite")
|
||||
return True
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error storing annotation in SQLite: {e}")
|
||||
import traceback
|
||||
traceback.print_exc()
|
||||
return False
|
||||
|
||||
def get_annotation(self, annotation_id: str) -> Optional[Dict[str, Any]]:
|
||||
"""
|
||||
Retrieve complete annotation with all market data
|
||||
|
||||
Args:
|
||||
annotation_id: Annotation ID
|
||||
|
||||
Returns:
|
||||
Dict with annotation data, OHLCV snapshots, and model predictions
|
||||
"""
|
||||
try:
|
||||
conn = sqlite3.connect(self.db_path)
|
||||
cursor = conn.cursor()
|
||||
|
||||
# Get annotation metadata
|
||||
cursor.execute("""
|
||||
SELECT * FROM annotations WHERE annotation_id = ?
|
||||
""", (annotation_id,))
|
||||
|
||||
row = cursor.fetchone()
|
||||
if not row:
|
||||
conn.close()
|
||||
return None
|
||||
|
||||
# Parse annotation data
|
||||
columns = [desc[0] for desc in cursor.description]
|
||||
annotation = dict(zip(columns, row))
|
||||
|
||||
# Parse JSON fields
|
||||
annotation['market_context'] = json.loads(annotation.get('market_context', '{}'))
|
||||
annotation['model_features'] = json.loads(annotation.get('model_features', '{}'))
|
||||
annotation['pivot_data'] = json.loads(annotation.get('pivot_data', '{}'))
|
||||
|
||||
# Get OHLCV snapshots
|
||||
cursor.execute("""
|
||||
SELECT timeframe, timestamp, open, high, low, close, volume
|
||||
FROM annotation_ohlcv_snapshots
|
||||
WHERE annotation_id = ?
|
||||
ORDER BY timeframe, timestamp
|
||||
""", (annotation_id,))
|
||||
|
||||
snapshots = {}
|
||||
for row in cursor.fetchall():
|
||||
timeframe = row[0]
|
||||
if timeframe not in snapshots:
|
||||
snapshots[timeframe] = []
|
||||
|
||||
snapshots[timeframe].append({
|
||||
'timestamp': row[1],
|
||||
'open': row[2],
|
||||
'high': row[3],
|
||||
'low': row[4],
|
||||
'close': row[5],
|
||||
'volume': row[6]
|
||||
})
|
||||
|
||||
# Convert to DataFrames
|
||||
annotation['ohlcv_snapshots'] = {}
|
||||
for timeframe, data in snapshots.items():
|
||||
df = pd.DataFrame(data)
|
||||
df['timestamp'] = pd.to_datetime(df['timestamp'], unit='ms', utc=True)
|
||||
df = df.set_index('timestamp')
|
||||
annotation['ohlcv_snapshots'][timeframe] = df
|
||||
|
||||
# Get model predictions
|
||||
cursor.execute("""
|
||||
SELECT model_name, timestamp, prediction, confidence, features
|
||||
FROM annotation_model_predictions
|
||||
WHERE annotation_id = ?
|
||||
""", (annotation_id,))
|
||||
|
||||
predictions = []
|
||||
for row in cursor.fetchall():
|
||||
predictions.append({
|
||||
'model_name': row[0],
|
||||
'timestamp': row[1],
|
||||
'prediction': json.loads(row[2]),
|
||||
'confidence': row[3],
|
||||
'features': json.loads(row[4])
|
||||
})
|
||||
|
||||
annotation['model_predictions'] = predictions
|
||||
|
||||
conn.close()
|
||||
|
||||
logger.info(f"Retrieved annotation {annotation_id} with {len(snapshots)} timeframes from SQLite")
|
||||
return annotation
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error retrieving annotation from SQLite: {e}")
|
||||
return None
|
||||
|
||||
def get_cache_stats(self) -> Dict[str, Any]:
|
||||
"""Get cache statistics"""
|
||||
try:
|
||||
conn = sqlite3.connect(self.db_path)
|
||||
cursor = conn.cursor()
|
||||
|
||||
# Get OHLCV stats
|
||||
cursor.execute("""
|
||||
SELECT symbol, timeframe, candle_count, first_timestamp, last_timestamp
|
||||
FROM cache_metadata
|
||||
ORDER BY symbol, timeframe
|
||||
""")
|
||||
|
||||
ohlcv_stats = []
|
||||
for row in cursor.fetchall():
|
||||
ohlcv_stats.append({
|
||||
'symbol': row[0],
|
||||
'timeframe': row[1],
|
||||
'candle_count': row[2],
|
||||
'first_timestamp': datetime.fromtimestamp(row[3] / 1000),
|
||||
'last_timestamp': datetime.fromtimestamp(row[4] / 1000)
|
||||
})
|
||||
|
||||
# Get annotation count
|
||||
cursor.execute("SELECT COUNT(*) FROM annotations")
|
||||
annotation_count = cursor.fetchone()[0]
|
||||
|
||||
# Get total OHLCV rows
|
||||
cursor.execute("SELECT COUNT(*) FROM ohlcv_data")
|
||||
total_candles = cursor.fetchone()[0]
|
||||
|
||||
conn.close()
|
||||
|
||||
return {
|
||||
'ohlcv_stats': ohlcv_stats,
|
||||
'annotation_count': annotation_count,
|
||||
'total_candles': total_candles
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error getting cache stats: {e}")
|
||||
return {}
|
||||
Reference in New Issue
Block a user