TCPDashboard/components/charts/layers/bot_integration.py
2025-06-12 13:27:30 +08:00

737 lines
28 KiB
Python

"""
Bot Management Integration for Chart Signal Layers
This module provides integration points between the signal layer system and the bot management
system, including data fetching utilities, bot filtering, and integration helpers.
"""
import pandas as pd
from typing import Dict, Any, Optional, List, Union, Tuple
from dataclasses import dataclass
from datetime import datetime, timedelta
from decimal import Decimal
from database.connection import get_session
from database.models import Bot, Signal, Trade, BotPerformance
from database.operations import DatabaseOperationError
from utils.logger import get_logger
# Initialize logger
logger = get_logger()
@dataclass
class BotFilterConfig:
"""Configuration for filtering bot data for chart layers"""
bot_ids: Optional[List[int]] = None # Specific bot IDs to include
bot_names: Optional[List[str]] = None # Specific bot names to include
strategies: Optional[List[str]] = None # Specific strategies to include
symbols: Optional[List[str]] = None # Specific symbols to include
statuses: Optional[List[str]] = None # Bot statuses to include
date_range: Optional[Tuple[datetime, datetime]] = None # Date range filter
active_only: bool = False # Only include active bots
def __post_init__(self):
if self.statuses is None:
self.statuses = ['active', 'inactive', 'paused'] # Exclude 'error' by default
class BotDataService:
"""
Service for fetching bot-related data for chart layers.
"""
def __init__(self):
"""Initialize bot data service."""
self.logger = logger
def get_bots(self, filter_config: BotFilterConfig = None) -> pd.DataFrame:
"""
Get bot information based on filter configuration.
Args:
filter_config: Filter configuration (optional)
Returns:
DataFrame with bot information
"""
try:
if filter_config is None:
filter_config = BotFilterConfig()
with get_session() as session:
query = session.query(Bot)
# Apply filters
if filter_config.bot_ids:
query = query.filter(Bot.id.in_(filter_config.bot_ids))
if filter_config.bot_names:
query = query.filter(Bot.name.in_(filter_config.bot_names))
if filter_config.strategies:
query = query.filter(Bot.strategy_name.in_(filter_config.strategies))
if filter_config.symbols:
query = query.filter(Bot.symbol.in_(filter_config.symbols))
if filter_config.statuses:
query = query.filter(Bot.status.in_(filter_config.statuses))
if filter_config.active_only:
query = query.filter(Bot.status == 'active')
# Execute query
bots = query.all()
# Convert to DataFrame
bot_data = []
for bot in bots:
bot_data.append({
'id': bot.id,
'name': bot.name,
'strategy_name': bot.strategy_name,
'symbol': bot.symbol,
'timeframe': bot.timeframe,
'status': bot.status,
'config_file': bot.config_file,
'virtual_balance': float(bot.virtual_balance) if bot.virtual_balance else 0.0,
'current_balance': float(bot.current_balance) if bot.current_balance else 0.0,
'pnl': float(bot.pnl) if bot.pnl else 0.0,
'is_active': bot.is_active,
'last_heartbeat': bot.last_heartbeat,
'created_at': bot.created_at,
'updated_at': bot.updated_at
})
df = pd.DataFrame(bot_data)
self.logger.info(f"Bot Integration: Retrieved {len(df)} bots with filters: {filter_config}")
return df
except Exception as e:
self.logger.error(f"Bot Integration: Error retrieving bots: {e}")
raise DatabaseOperationError(f"Failed to retrieve bots: {e}")
def get_signals_for_bots(self,
bot_ids: Union[int, List[int]] = None,
start_time: datetime = None,
end_time: datetime = None,
signal_types: List[str] = None,
min_confidence: float = 0.0) -> pd.DataFrame:
"""
Get signals for specific bots or all bots.
Args:
bot_ids: Bot ID(s) to fetch signals for (None for all bots)
start_time: Start time for signal filtering
end_time: End time for signal filtering
signal_types: Signal types to include (['buy', 'sell', 'hold'])
min_confidence: Minimum confidence threshold
Returns:
DataFrame with signal data
"""
try:
# Default time range if not provided
if end_time is None:
end_time = datetime.now()
if start_time is None:
start_time = end_time - timedelta(days=7) # Last 7 days by default
# Normalize bot_ids to list
if isinstance(bot_ids, int):
bot_ids = [bot_ids]
with get_session() as session:
query = session.query(Signal)
# Apply filters
if bot_ids is not None:
query = query.filter(Signal.bot_id.in_(bot_ids))
query = query.filter(
Signal.timestamp >= start_time,
Signal.timestamp <= end_time
)
if signal_types:
query = query.filter(Signal.signal_type.in_(signal_types))
if min_confidence > 0:
query = query.filter(Signal.confidence >= min_confidence)
# Order by timestamp
query = query.order_by(Signal.timestamp.asc())
# Execute query
signals = query.all()
# Convert to DataFrame
signal_data = []
for signal in signals:
signal_data.append({
'id': signal.id,
'bot_id': signal.bot_id,
'timestamp': signal.timestamp,
'signal_type': signal.signal_type,
'price': float(signal.price) if signal.price else None,
'confidence': float(signal.confidence) if signal.confidence else None,
'indicators': signal.indicators, # JSONB data
'created_at': signal.created_at
})
df = pd.DataFrame(signal_data)
self.logger.info(f"Bot Integration: Retrieved {len(df)} signals for bots: {bot_ids}")
return df
except Exception as e:
self.logger.error(f"Bot Integration: Error retrieving signals: {e}")
raise DatabaseOperationError(f"Failed to retrieve signals: {e}")
def get_trades_for_bots(self,
bot_ids: Union[int, List[int]] = None,
start_time: datetime = None,
end_time: datetime = None,
sides: List[str] = None) -> pd.DataFrame:
"""
Get trades for specific bots or all bots.
Args:
bot_ids: Bot ID(s) to fetch trades for (None for all bots)
start_time: Start time for trade filtering
end_time: End time for trade filtering
sides: Trade sides to include (['buy', 'sell'])
Returns:
DataFrame with trade data
"""
try:
# Default time range if not provided
if end_time is None:
end_time = datetime.now()
if start_time is None:
start_time = end_time - timedelta(days=7) # Last 7 days by default
# Normalize bot_ids to list
if isinstance(bot_ids, int):
bot_ids = [bot_ids]
with get_session() as session:
query = session.query(Trade)
# Apply filters
if bot_ids is not None:
query = query.filter(Trade.bot_id.in_(bot_ids))
query = query.filter(
Trade.timestamp >= start_time,
Trade.timestamp <= end_time
)
if sides:
query = query.filter(Trade.side.in_(sides))
# Order by timestamp
query = query.order_by(Trade.timestamp.asc())
# Execute query
trades = query.all()
# Convert to DataFrame
trade_data = []
for trade in trades:
trade_data.append({
'id': trade.id,
'bot_id': trade.bot_id,
'signal_id': trade.signal_id,
'timestamp': trade.timestamp,
'side': trade.side,
'price': float(trade.price),
'quantity': float(trade.quantity),
'fees': float(trade.fees),
'pnl': float(trade.pnl) if trade.pnl else None,
'balance_after': float(trade.balance_after) if trade.balance_after else None,
'trade_value': float(trade.trade_value),
'net_pnl': float(trade.net_pnl),
'created_at': trade.created_at
})
df = pd.DataFrame(trade_data)
self.logger.info(f"Bot Integration: Retrieved {len(df)} trades for bots: {bot_ids}")
return df
except Exception as e:
self.logger.error(f"Bot Integration: Error retrieving trades: {e}")
raise DatabaseOperationError(f"Failed to retrieve trades: {e}")
def get_bot_performance(self,
bot_ids: Union[int, List[int]] = None,
start_time: datetime = None,
end_time: datetime = None) -> pd.DataFrame:
"""
Get performance data for specific bots.
Args:
bot_ids: Bot ID(s) to fetch performance for (None for all bots)
start_time: Start time for performance filtering
end_time: End time for performance filtering
Returns:
DataFrame with performance data
"""
try:
# Default time range if not provided
if end_time is None:
end_time = datetime.now()
if start_time is None:
start_time = end_time - timedelta(days=30) # Last 30 days by default
# Normalize bot_ids to list
if isinstance(bot_ids, int):
bot_ids = [bot_ids]
with get_session() as session:
query = session.query(BotPerformance)
# Apply filters
if bot_ids is not None:
query = query.filter(BotPerformance.bot_id.in_(bot_ids))
query = query.filter(
BotPerformance.timestamp >= start_time,
BotPerformance.timestamp <= end_time
)
# Order by timestamp
query = query.order_by(BotPerformance.timestamp.asc())
# Execute query
performance_records = query.all()
# Convert to DataFrame
performance_data = []
for perf in performance_records:
performance_data.append({
'id': perf.id,
'bot_id': perf.bot_id,
'timestamp': perf.timestamp,
'total_value': float(perf.total_value),
'cash_balance': float(perf.cash_balance),
'crypto_balance': float(perf.crypto_balance),
'total_trades': perf.total_trades,
'winning_trades': perf.winning_trades,
'total_fees': float(perf.total_fees),
'win_rate': perf.win_rate,
'portfolio_allocation': perf.portfolio_allocation,
'created_at': perf.created_at
})
df = pd.DataFrame(performance_data)
self.logger.info(f"Bot Integration: Retrieved {len(df)} performance records for bots: {bot_ids}")
return df
except Exception as e:
self.logger.error(f"Bot Integration: Error retrieving bot performance: {e}")
raise DatabaseOperationError(f"Failed to retrieve bot performance: {e}")
class BotSignalLayerIntegration:
"""
Integration utilities for signal layers with bot management system.
"""
def __init__(self):
"""Initialize bot signal layer integration."""
self.data_service = BotDataService()
self.logger = logger
def get_signals_for_chart(self,
symbol: str,
timeframe: str = None,
bot_filter: BotFilterConfig = None,
time_range: Tuple[datetime, datetime] = None,
signal_types: List[str] = None,
min_confidence: float = 0.0) -> pd.DataFrame:
"""
Get signals filtered by chart context (symbol, timeframe) and bot criteria.
Args:
symbol: Trading symbol for the chart
timeframe: Chart timeframe (optional)
bot_filter: Bot filtering configuration
time_range: (start_time, end_time) tuple
signal_types: Signal types to include
min_confidence: Minimum confidence threshold
Returns:
DataFrame with signals ready for chart rendering
"""
try:
# Get relevant bots for this symbol/timeframe
if bot_filter is None:
bot_filter = BotFilterConfig()
# Add symbol filter
if bot_filter.symbols is None:
bot_filter.symbols = [symbol]
elif symbol not in bot_filter.symbols:
bot_filter.symbols.append(symbol)
# Get bots matching criteria
bots_df = self.data_service.get_bots(bot_filter)
if bots_df.empty:
self.logger.info(f"No bots found for symbol {symbol}")
return pd.DataFrame()
bot_ids = bots_df['id'].tolist()
# Get time range
start_time, end_time = time_range if time_range else (None, None)
# Get signals for these bots
signals_df = self.data_service.get_signals_for_bots(
bot_ids=bot_ids,
start_time=start_time,
end_time=end_time,
signal_types=signal_types,
min_confidence=min_confidence
)
# Enrich signals with bot information
if not signals_df.empty:
signals_df = signals_df.merge(
bots_df[['id', 'name', 'strategy_name', 'status']],
left_on='bot_id',
right_on='id',
suffixes=('', '_bot')
)
# Add metadata fields for chart rendering
signals_df['bot_name'] = signals_df['name']
signals_df['strategy'] = signals_df['strategy_name']
signals_df['bot_status'] = signals_df['status']
# Clean up duplicate columns
signals_df = signals_df.drop(['id_bot', 'name', 'strategy_name', 'status'], axis=1)
self.logger.info(f"Bot Integration: Retrieved {len(signals_df)} signals for chart {symbol} from {len(bot_ids)} bots")
return signals_df
except Exception as e:
self.logger.error(f"Bot Integration: Error getting signals for chart: {e}")
return pd.DataFrame()
def get_trades_for_chart(self,
symbol: str,
timeframe: str = None,
bot_filter: BotFilterConfig = None,
time_range: Tuple[datetime, datetime] = None,
sides: List[str] = None) -> pd.DataFrame:
"""
Get trades filtered by chart context (symbol, timeframe) and bot criteria.
Args:
symbol: Trading symbol for the chart
timeframe: Chart timeframe (optional)
bot_filter: Bot filtering configuration
time_range: (start_time, end_time) tuple
sides: Trade sides to include
Returns:
DataFrame with trades ready for chart rendering
"""
try:
# Get relevant bots for this symbol/timeframe
if bot_filter is None:
bot_filter = BotFilterConfig()
# Add symbol filter
if bot_filter.symbols is None:
bot_filter.symbols = [symbol]
elif symbol not in bot_filter.symbols:
bot_filter.symbols.append(symbol)
# Get bots matching criteria
bots_df = self.data_service.get_bots(bot_filter)
if bots_df.empty:
self.logger.info(f"No bots found for symbol {symbol}")
return pd.DataFrame()
bot_ids = bots_df['id'].tolist()
# Get time range
start_time, end_time = time_range if time_range else (None, None)
# Get trades for these bots
trades_df = self.data_service.get_trades_for_bots(
bot_ids=bot_ids,
start_time=start_time,
end_time=end_time,
sides=sides
)
# Enrich trades with bot information
if not trades_df.empty:
trades_df = trades_df.merge(
bots_df[['id', 'name', 'strategy_name', 'status']],
left_on='bot_id',
right_on='id',
suffixes=('', '_bot')
)
# Add metadata fields for chart rendering
trades_df['bot_name'] = trades_df['name']
trades_df['strategy'] = trades_df['strategy_name']
trades_df['bot_status'] = trades_df['status']
# Clean up duplicate columns
trades_df = trades_df.drop(['id_bot', 'name', 'strategy_name', 'status'], axis=1)
self.logger.info(f"Bot Integration: Retrieved {len(trades_df)} trades for chart {symbol} from {len(bot_ids)} bots")
return trades_df
except Exception as e:
self.logger.error(f"Bot Integration: Error getting trades for chart: {e}")
return pd.DataFrame()
def get_bot_summary_stats(self, bot_ids: List[int] = None) -> Dict[str, Any]:
"""
Get summary statistics for bots.
Args:
bot_ids: Specific bot IDs (None for all bots)
Returns:
Dictionary with summary statistics
"""
try:
# Get bots
bot_filter = BotFilterConfig(bot_ids=bot_ids) if bot_ids else BotFilterConfig()
bots_df = self.data_service.get_bots(bot_filter)
if bots_df.empty:
return {
'total_bots': 0,
'active_bots': 0,
'total_balance': 0.0,
'total_pnl': 0.0,
'strategies': [],
'symbols': []
}
# Calculate statistics
stats = {
'total_bots': len(bots_df),
'active_bots': len(bots_df[bots_df['status'] == 'active']),
'inactive_bots': len(bots_df[bots_df['status'] == 'inactive']),
'paused_bots': len(bots_df[bots_df['status'] == 'paused']),
'error_bots': len(bots_df[bots_df['status'] == 'error']),
'total_virtual_balance': bots_df['virtual_balance'].sum(),
'total_current_balance': bots_df['current_balance'].sum(),
'total_pnl': bots_df['pnl'].sum(),
'average_pnl': bots_df['pnl'].mean(),
'best_performing_bot': None,
'worst_performing_bot': None,
'strategies': bots_df['strategy_name'].unique().tolist(),
'symbols': bots_df['symbol'].unique().tolist(),
'timeframes': bots_df['timeframe'].unique().tolist()
}
# Get best and worst performing bots
if not bots_df.empty:
best_bot = bots_df.loc[bots_df['pnl'].idxmax()]
worst_bot = bots_df.loc[bots_df['pnl'].idxmin()]
stats['best_performing_bot'] = {
'id': best_bot['id'],
'name': best_bot['name'],
'pnl': best_bot['pnl']
}
stats['worst_performing_bot'] = {
'id': worst_bot['id'],
'name': worst_bot['name'],
'pnl': worst_bot['pnl']
}
return stats
except Exception as e:
self.logger.error(f"Bot Integration: Error getting bot summary stats: {e}")
return {}
# Global instances for easy access
bot_data_service = BotDataService()
bot_integration = BotSignalLayerIntegration()
# Convenience functions for common use cases
def get_active_bot_signals(symbol: str,
timeframe: str = None,
days_back: int = 7,
signal_types: List[str] = None,
min_confidence: float = 0.3) -> pd.DataFrame:
"""
Get signals from active bots for a specific symbol.
Args:
symbol: Trading symbol
timeframe: Chart timeframe (optional)
days_back: Number of days to look back
signal_types: Signal types to include
min_confidence: Minimum confidence threshold
Returns:
DataFrame with signals from active bots
"""
end_time = datetime.now()
start_time = end_time - timedelta(days=days_back)
bot_filter = BotFilterConfig(
symbols=[symbol],
active_only=True
)
return bot_integration.get_signals_for_chart(
symbol=symbol,
timeframe=timeframe,
bot_filter=bot_filter,
time_range=(start_time, end_time),
signal_types=signal_types,
min_confidence=min_confidence
)
def get_active_bot_trades(symbol: str,
timeframe: str = None,
days_back: int = 7,
sides: List[str] = None) -> pd.DataFrame:
"""
Get trades from active bots for a specific symbol.
Args:
symbol: Trading symbol
timeframe: Chart timeframe (optional)
days_back: Number of days to look back
sides: Trade sides to include
Returns:
DataFrame with trades from active bots
"""
end_time = datetime.now()
start_time = end_time - timedelta(days=days_back)
bot_filter = BotFilterConfig(
symbols=[symbol],
active_only=True
)
return bot_integration.get_trades_for_chart(
symbol=symbol,
timeframe=timeframe,
bot_filter=bot_filter,
time_range=(start_time, end_time),
sides=sides
)
def get_bot_signals_by_strategy(strategy_name: str,
symbol: str = None,
days_back: int = 7,
signal_types: List[str] = None) -> pd.DataFrame:
"""
Get signals from bots using a specific strategy.
Args:
strategy_name: Strategy name to filter by
symbol: Trading symbol (optional)
days_back: Number of days to look back
signal_types: Signal types to include
Returns:
DataFrame with signals from strategy bots
"""
end_time = datetime.now()
start_time = end_time - timedelta(days=days_back)
bot_filter = BotFilterConfig(
strategies=[strategy_name],
symbols=[symbol] if symbol else None
)
# Get bots for this strategy
bots_df = bot_data_service.get_bots(bot_filter)
if bots_df.empty:
return pd.DataFrame()
bot_ids = bots_df['id'].tolist()
return bot_data_service.get_signals_for_bots(
bot_ids=bot_ids,
start_time=start_time,
end_time=end_time,
signal_types=signal_types
)
def get_bot_performance_summary(bot_id: int = None,
days_back: int = 30) -> Dict[str, Any]:
"""
Get performance summary for a specific bot or all bots.
Args:
bot_id: Specific bot ID (None for all bots)
days_back: Number of days to analyze
Returns:
Dictionary with performance summary
"""
end_time = datetime.now()
start_time = end_time - timedelta(days=days_back)
# Get bot summary stats
bot_ids = [bot_id] if bot_id else None
bot_stats = bot_integration.get_bot_summary_stats(bot_ids)
# Get signals and trades for performance analysis
signals_df = bot_data_service.get_signals_for_bots(
bot_ids=bot_ids,
start_time=start_time,
end_time=end_time
)
trades_df = bot_data_service.get_trades_for_bots(
bot_ids=bot_ids,
start_time=start_time,
end_time=end_time
)
# Calculate additional performance metrics
performance = {
'bot_stats': bot_stats,
'signal_count': len(signals_df),
'trade_count': len(trades_df),
'signals_by_type': signals_df['signal_type'].value_counts().to_dict() if not signals_df.empty else {},
'trades_by_side': trades_df['side'].value_counts().to_dict() if not trades_df.empty else {},
'total_trade_volume': trades_df['trade_value'].sum() if not trades_df.empty else 0.0,
'total_fees': trades_df['fees'].sum() if not trades_df.empty else 0.0,
'profitable_trades': len(trades_df[trades_df['pnl'] > 0]) if not trades_df.empty else 0,
'losing_trades': len(trades_df[trades_df['pnl'] < 0]) if not trades_df.empty else 0,
'win_rate': (len(trades_df[trades_df['pnl'] > 0]) / len(trades_df) * 100) if not trades_df.empty else 0.0,
'time_range': {
'start': start_time.isoformat(),
'end': end_time.isoformat(),
'days': days_back
}
}
return performance