File size: 9,012 Bytes
9cb5a00 | 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 | """
ML-3m-trader Metrics & Evaluation
====================================
Computes comprehensive trading performance metrics and generates
a formatted report.
"""
import os
from datetime import timedelta
import numpy as np
import pandas as pd
import config as cfg
def compute_metrics(
trades: list,
equity_curve: np.ndarray,
starting_balance: float = cfg.STARTING_BALANCE,
) -> dict:
"""
Compute all performance metrics from backtest results.
Parameters
----------
trades : list of dict
Trade records from backtester.
equity_curve : np.ndarray
Balance at each bar.
starting_balance : float
Initial account balance.
Returns
-------
dict
All computed metrics.
"""
if not trades:
return {"error": "No trades executed."}
df_trades = pd.DataFrame(trades)
# Basic counts
total_trades = len(df_trades)
wins = df_trades[df_trades["result"] == "TP"]
losses = df_trades[df_trades["result"] == "SL"]
n_wins = len(wins)
n_losses = len(losses)
win_rate = n_wins / total_trades if total_trades > 0 else 0.0
# PnL
pnl_array = df_trades["pnl"].values
gross_profit = pnl_array[pnl_array > 0].sum() if (pnl_array > 0).any() else 0.0
gross_loss = abs(pnl_array[pnl_array < 0].sum()) if (pnl_array < 0).any() else 0.0
net_profit = pnl_array.sum()
# Average win/loss as percentage of balance at entry
avg_win_pct = 0.0
avg_loss_pct = 0.0
if n_wins > 0:
# Each trade's PnL as % of balance before trade
win_pcts = []
for _, t in wins.iterrows():
bal_before = t["balance_after"] - t["pnl"]
if bal_before > 0:
win_pcts.append(t["pnl"] / bal_before * 100.0)
avg_win_pct = np.mean(win_pcts) if win_pcts else 0.0
if n_losses > 0:
loss_pcts = []
for _, t in losses.iterrows():
bal_before = t["balance_after"] - t["pnl"]
if bal_before > 0:
loss_pcts.append(abs(t["pnl"]) / bal_before * 100.0)
avg_loss_pct = np.mean(loss_pcts) if loss_pcts else 0.0
# Profit factor
profit_factor = gross_profit / gross_loss if gross_loss > 0 else float("inf")
# Expectancy
expectancy = net_profit / total_trades if total_trades > 0 else 0.0
# Equity curve metrics
final_balance = equity_curve[-1]
total_return = (final_balance - starting_balance) / starting_balance * 100.0
# Max drawdown
running_max = np.maximum.accumulate(equity_curve)
drawdowns = (equity_curve - running_max) / running_max
max_drawdown = abs(drawdowns.min()) * 100.0 # percentage
# Sharpe Ratio (annualized from bar-level returns)
eq_returns = np.diff(equity_curve) / equity_curve[:-1]
eq_returns = eq_returns[np.isfinite(eq_returns)]
bars_per_day = (6.5 * 60) / cfg.TIMEFRAME_MINUTES
annual_factor = np.sqrt(252 * bars_per_day)
if len(eq_returns) > 1 and np.std(eq_returns) > 0:
sharpe = (np.mean(eq_returns) / np.std(eq_returns)) * annual_factor
else:
sharpe = 0.0
# Sortino Ratio
downside = eq_returns[eq_returns < 0]
if len(downside) > 1 and np.std(downside) > 0:
sortino = (np.mean(eq_returns) / np.std(downside)) * annual_factor
else:
sortino = 0.0
# Calmar Ratio
# Annualized return / max drawdown
total_bars = len(equity_curve)
years = total_bars / (bars_per_day * 252) if bars_per_day > 0 else 1.0
annualized_return = ((final_balance / starting_balance) ** (1 / max(years, 0.01)) - 1) * 100.0
calmar = annualized_return / max_drawdown if max_drawdown > 0 else float("inf")
# Trade duration
durations = (df_trades["exit_bar"] - df_trades["entry_bar"]).values
avg_duration_bars = np.mean(durations) if len(durations) > 0 else 0
avg_duration_min = avg_duration_bars * cfg.TIMEFRAME_MINUTES
# Intraday PnL stats
daily_pnl = {}
if "time_entry" in df_trades.columns:
for _, t in df_trades.iterrows():
if t["time_entry"] is not None:
day = pd.Timestamp(t["time_entry"]).date()
daily_pnl.setdefault(day, 0.0)
daily_pnl[day] += t["pnl"]
daily_pnl_values = np.array(list(daily_pnl.values())) if daily_pnl else np.array([0.0])
daily_mean = np.mean(daily_pnl_values)
daily_std = np.std(daily_pnl_values) if len(daily_pnl_values) > 1 else 0.0
daily_min = np.min(daily_pnl_values)
daily_max = np.max(daily_pnl_values)
# Buy/Sell breakdown
buy_trades = df_trades[df_trades["direction"] == "BUY"]
sell_trades = df_trades[df_trades["direction"] == "SELL"]
metrics = {
"total_trades": total_trades,
"winning_trades": n_wins,
"losing_trades": n_losses,
"win_rate": win_rate,
"avg_win_pct": avg_win_pct,
"avg_loss_pct": avg_loss_pct,
"gross_profit": gross_profit,
"gross_loss": gross_loss,
"net_profit": net_profit,
"profit_factor": profit_factor,
"expectancy": expectancy,
"starting_balance": starting_balance,
"final_balance": final_balance,
"total_return_pct": total_return,
"max_drawdown_pct": max_drawdown,
"sharpe_ratio": sharpe,
"sortino_ratio": sortino,
"calmar_ratio": calmar,
"avg_trade_duration_bars": avg_duration_bars,
"avg_trade_duration_min": avg_duration_min,
"total_buy_trades": len(buy_trades),
"total_sell_trades": len(sell_trades),
"daily_pnl_mean": daily_mean,
"daily_pnl_std": daily_std,
"daily_pnl_min": daily_min,
"daily_pnl_max": daily_max,
"trading_days": len(daily_pnl),
}
return metrics
def format_report(metrics: dict) -> str:
"""Format metrics into a readable console report."""
if "error" in metrics:
return f"\n[REPORT] {metrics['error']}\n"
sep = "=" * 60
lines = [
"",
sep,
" ML-3m-trader BACKTEST REPORT",
sep,
"",
" ACCOUNT",
f" Starting Balance : ${metrics['starting_balance']:>12,.2f}",
f" Final Balance : ${metrics['final_balance']:>12,.2f}",
f" Net Profit : ${metrics['net_profit']:>12,.2f}",
f" Total Return : {metrics['total_return_pct']:>11.2f}%",
"",
" TRADE STATISTICS",
f" Total Trades : {metrics['total_trades']:>12,}",
f" Winning Trades : {metrics['winning_trades']:>12,}",
f" Losing Trades : {metrics['losing_trades']:>12,}",
f" Win Rate : {metrics['win_rate'] * 100:>11.2f}%",
f" Buy Trades : {metrics['total_buy_trades']:>12,}",
f" Sell Trades : {metrics['total_sell_trades']:>12,}",
"",
" PROFIT / LOSS",
f" Avg Win : {metrics['avg_win_pct']:>11.4f}%",
f" Avg Loss : {metrics['avg_loss_pct']:>11.4f}%",
f" Gross Profit : ${metrics['gross_profit']:>12,.2f}",
f" Gross Loss : ${metrics['gross_loss']:>12,.2f}",
f" Profit Factor : {metrics['profit_factor']:>12.4f}",
f" Expectancy (per trade): ${metrics['expectancy']:>12,.2f}",
"",
" RISK METRICS",
f" Max Drawdown : {metrics['max_drawdown_pct']:>11.2f}%",
f" Sharpe Ratio : {metrics['sharpe_ratio']:>12.4f}",
f" Sortino Ratio : {metrics['sortino_ratio']:>12.4f}",
f" Calmar Ratio : {metrics['calmar_ratio']:>12.4f}",
"",
" DURATION",
f" Avg Duration (bars) : {metrics['avg_trade_duration_bars']:>12.1f}",
f" Avg Duration (min) : {metrics['avg_trade_duration_min']:>12.1f}",
"",
" INTRADAY PNL",
f" Trading Days : {metrics['trading_days']:>12,}",
f" Daily Mean PnL : ${metrics['daily_pnl_mean']:>12,.2f}",
f" Daily Std PnL : ${metrics['daily_pnl_std']:>12,.2f}",
f" Daily Min PnL : ${metrics['daily_pnl_min']:>12,.2f}",
f" Daily Max PnL : ${metrics['daily_pnl_max']:>12,.2f}",
"",
sep,
]
return "\n".join(lines)
def save_report(report: str, filename: str = "report.txt"):
"""Save the report to the results directory."""
os.makedirs(cfg.RESULTS_DIR, exist_ok=True)
path = os.path.join(cfg.RESULTS_DIR, filename)
with open(path, "w", encoding="utf-8") as f:
f.write(report)
print(f"[INFO] Report saved to {path}")
def save_trades_csv(trades: list, filename: str = "trades.csv"):
"""Save individual trade records to CSV."""
os.makedirs(cfg.RESULTS_DIR, exist_ok=True)
path = os.path.join(cfg.RESULTS_DIR, filename)
pd.DataFrame(trades).to_csv(path, index=False)
print(f"[INFO] Trade log saved to {path}")
|