768 lines
28 KiB
Python
768 lines
28 KiB
Python
"""
|
||
量化交易回测系统 - EMA交叉策略（双EMA/三EMA）
|
||
"""
|
||
import csv
|
||
import datetime
|
||
import numpy as np
|
||
from typing import List, Dict, Optional, Tuple
|
||
|
||
|
||
# ========================= EMA计算函数 =========================
|
||
|
||
def calculate_ema(prices: List[float], period: int) -> List[Optional[float]]:
|
||
"""
|
||
计算指数移动平均线(EMA)
|
||
|
||
Args:
|
||
prices: 价格列表（通常是收盘价）
|
||
period: EMA周期
|
||
|
||
Returns:
|
||
EMA值列表，前period-1个为None
|
||
"""
|
||
if len(prices) < period:
|
||
return [None] * len(prices)
|
||
|
||
ema_values = [None] * (period - 1)
|
||
|
||
# 计算初始SMA作为EMA的起点
|
||
sma = sum(prices[:period]) / period
|
||
|
||
# EMA计算公式：EMA_today = (Price_today * (2/(period+1))) + (EMA_yesterday * (1 - (2/(period+1))))
|
||
multiplier = 2 / (period + 1)
|
||
|
||
# 第一个EMA值
|
||
ema = sma
|
||
ema_values.append(ema)
|
||
|
||
# 计算后续EMA值
|
||
for price in prices[period:]:
|
||
ema = (price * multiplier) + (ema * (1 - multiplier))
|
||
ema_values.append(ema)
|
||
|
||
return ema_values
|
||
|
||
|
||
# ========================= 策略核心函数 =========================
|
||
|
||
def check_ema_cross(ema_fast: List[Optional[float]],
|
||
ema_slow: List[Optional[float]],
|
||
idx: int) -> Optional[str]:
|
||
"""
|
||
检查EMA金叉/死叉
|
||
|
||
Args:
|
||
ema_fast: 快线EMA值列表
|
||
ema_slow: 慢线EMA值列表
|
||
idx: 当前K线索引
|
||
|
||
Returns:
|
||
"golden" - 金叉（做多信号）
|
||
"dead" - 死叉（做空信号）
|
||
None - 无交叉
|
||
"""
|
||
if idx < 1:
|
||
return None
|
||
|
||
# 确保有足够的数据
|
||
if ema_fast[idx] is None or ema_fast[idx - 1] is None:
|
||
return None
|
||
if ema_slow[idx] is None or ema_slow[idx - 1] is None:
|
||
return None
|
||
|
||
# 前一根K线快线在慢线下方，当前K线快线上穿慢线 -> 金叉
|
||
if ema_fast[idx - 1] < ema_slow[idx - 1] and ema_fast[idx] > ema_slow[idx]:
|
||
return "golden"
|
||
|
||
# 前一根K线快线在慢线上方，当前K线快线下穿慢线 -> 死叉
|
||
if ema_fast[idx - 1] > ema_slow[idx - 1] and ema_fast[idx] < ema_slow[idx]:
|
||
return "dead"
|
||
|
||
return None
|
||
|
||
|
||
def check_triple_ema_cross(ema_fast: List[Optional[float]],
|
||
ema_mid: List[Optional[float]],
|
||
ema_slow: List[Optional[float]],
|
||
idx: int) -> Optional[str]:
|
||
"""
|
||
检查三EMA交叉（更稳定的信号）
|
||
规则：快线 > 中线 > 慢线 -> 多头排列 -> 做多
|
||
快线 < 中线 < 慢线 -> 空头排列 -> 做空
|
||
|
||
Args:
|
||
ema_fast: 快线（如EMA7）
|
||
ema_mid: 中线（如EMA14）
|
||
ema_slow: 慢线（如EMA30）
|
||
idx: 当前K线索引
|
||
|
||
Returns:
|
||
"golden" - 金叉多头排列
|
||
"dead" - 死叉空头排列
|
||
None - 无明确信号
|
||
"""
|
||
if idx < 1:
|
||
return None
|
||
|
||
# 确保有足够的数据
|
||
if any(ema[idx] is None or ema[idx - 1] is None for ema in [ema_fast, ema_mid, ema_slow]):
|
||
return None
|
||
|
||
# 检查是否形成多头排列（EMA快 > 中 > 慢）
|
||
current_fast = ema_fast[idx]
|
||
current_mid = ema_mid[idx]
|
||
current_slow = ema_slow[idx]
|
||
prev_fast = ema_fast[idx - 1]
|
||
prev_mid = ema_mid[idx - 1]
|
||
prev_slow = ema_slow[idx - 1]
|
||
|
||
# 多头排列条件：快线 > 中线 > 慢线
|
||
is_golden_triple = current_fast > current_mid > current_slow
|
||
was_not_golden = not (prev_fast > prev_mid > prev_slow)
|
||
|
||
# 空头排列条件：快线 < 中线 < 慢线
|
||
is_dead_triple = current_fast < current_mid < current_slow
|
||
was_not_dead = not (prev_fast < prev_mid < prev_slow)
|
||
|
||
if is_golden_triple and was_not_golden:
|
||
return "golden"
|
||
elif is_dead_triple and was_not_dead:
|
||
return "dead"
|
||
|
||
return None
|
||
|
||
|
||
# ========================= 回测引擎 =========================
|
||
|
||
class EMABacktester:
|
||
"""EMA交叉策略回测器"""
|
||
|
||
def __init__(self,
|
||
fast_period: int = 12,
|
||
slow_period: int = 26,
|
||
signal_period: int = 9, # 用于MACD信号线
|
||
use_triple_ema: bool = False,
|
||
mid_period: int = 14, # 三EMA时的中间周期
|
||
use_macd_confirmation: bool = False,
|
||
stop_loss_pct: float = 0.02, # 2%止损
|
||
take_profit_pct: float = 0.05, # 5%止盈
|
||
trailing_stop_pct: float = 0.03): # 3%移动止损
|
||
|
||
self.fast_period = fast_period
|
||
self.slow_period = slow_period
|
||
self.signal_period = signal_period
|
||
self.use_triple_ema = use_triple_ema
|
||
self.mid_period = mid_period
|
||
self.use_macd_confirmation = use_macd_confirmation
|
||
self.stop_loss_pct = stop_loss_pct
|
||
self.take_profit_pct = take_profit_pct
|
||
self.trailing_stop_pct = trailing_stop_pct
|
||
|
||
self.stats = {
|
||
'golden_cross': {'count': 0, 'wins': 0, 'total_profit': 0.0, 'name': '金叉做多'},
|
||
'dead_cross': {'count': 0, 'wins': 0, 'total_profit': 0.0, 'name': '死叉做空'},
|
||
}
|
||
|
||
def calculate_macd(self, prices: List[float]) -> Tuple[List[Optional[float]],
|
||
List[Optional[float]],
|
||
List[Optional[float]]]:
|
||
"""
|
||
计算MACD指标
|
||
Returns: (MACD线, 信号线, 柱状图)
|
||
"""
|
||
# 计算快慢EMA
|
||
ema_fast = calculate_ema(prices, self.fast_period)
|
||
ema_slow = calculate_ema(prices, self.slow_period)
|
||
|
||
# 计算MACD线 = EMA快线 - EMA慢线
|
||
macd_line = []
|
||
for i in range(len(prices)):
|
||
if ema_fast[i] is not None and ema_slow[i] is not None:
|
||
macd_line.append(ema_fast[i] - ema_slow[i])
|
||
else:
|
||
macd_line.append(None)
|
||
|
||
# 计算信号线（MACD的EMA）
|
||
signal_line = calculate_ema([x for x in macd_line if x is not None] if any(macd_line) else [],
|
||
self.signal_period)
|
||
|
||
# 补全None值
|
||
signal_line_extended = [None] * (len(prices) - len(signal_line)) + signal_line if len(signal_line) < len(
|
||
prices) else signal_line
|
||
|
||
# 计算柱状图
|
||
histogram = []
|
||
for i in range(len(prices)):
|
||
if macd_line[i] is not None and signal_line_extended[i] is not None:
|
||
histogram.append(macd_line[i] - signal_line_extended[i])
|
||
else:
|
||
histogram.append(None)
|
||
|
||
return macd_line, signal_line_extended, histogram
|
||
|
||
def backtest(self, data: List[Dict]) -> Tuple[List[Dict], Dict]:
|
||
"""
|
||
执行EMA交叉策略回测
|
||
|
||
Args:
|
||
data: K线数据列表
|
||
|
||
Returns:
|
||
trades: 交易记录列表
|
||
stats: 统计数据
|
||
"""
|
||
# 提取收盘价
|
||
close_prices = [float(c['close']) for c in data]
|
||
|
||
# 计算EMA
|
||
ema_fast = calculate_ema(close_prices, self.fast_period)
|
||
ema_slow = calculate_ema(close_prices, self.slow_period)
|
||
|
||
# 如果需要三EMA，计算中间EMA
|
||
ema_mid = None
|
||
if self.use_triple_ema:
|
||
ema_mid = calculate_ema(close_prices, self.mid_period)
|
||
|
||
# 如果需要MACD确认，计算MACD
|
||
macd_line, signal_line, histogram = None, None, None
|
||
if self.use_macd_confirmation:
|
||
macd_line, signal_line, histogram = self.calculate_macd(close_prices)
|
||
|
||
trades: List[Dict] = []
|
||
current_position: Optional[Dict] = None
|
||
highest_price_since_entry = 0 # 用于移动止损
|
||
lowest_price_since_entry = float('inf') # 用于移动止损
|
||
|
||
# 遍历K线数据（跳过前几个没有EMA值的）
|
||
start_idx = max(self.fast_period, self.slow_period)
|
||
if self.use_triple_ema:
|
||
start_idx = max(start_idx, self.mid_period)
|
||
|
||
for idx in range(start_idx, len(data)):
|
||
current_bar = data[idx]
|
||
current_price = float(current_bar['close'])
|
||
open_price = float(current_bar['open'])
|
||
|
||
# ========== 信号检测 ==========
|
||
signal = None
|
||
|
||
# 基础双EMA交叉信号
|
||
if not self.use_triple_ema:
|
||
signal = check_ema_cross(ema_fast, ema_slow, idx)
|
||
# 三EMA排列信号
|
||
else:
|
||
signal = check_triple_ema_cross(ema_fast, ema_mid, ema_slow, idx)
|
||
|
||
# MACD确认（可选）
|
||
if signal and self.use_macd_confirmation:
|
||
macd_confirmed = False
|
||
if signal == "golden":
|
||
# 金叉确认：MACD线在信号线上方且柱状图为正
|
||
if macd_line[idx] is not None and signal_line[idx] is not None:
|
||
macd_confirmed = (macd_line[idx] > signal_line[idx]) and (
|
||
histogram[idx] is not None and histogram[idx] > 0)
|
||
elif signal == "dead":
|
||
# 死叉确认：MACD线在信号线下方且柱状图为负
|
||
if macd_line[idx] is not None and signal_line[idx] is not None:
|
||
macd_confirmed = (macd_line[idx] < signal_line[idx]) and (
|
||
histogram[idx] is not None and histogram[idx] < 0)
|
||
|
||
if not macd_confirmed:
|
||
signal = None
|
||
|
||
# ========== 空仓时开仓 ==========
|
||
if current_position is None and signal:
|
||
# 下一根K线开盘价入场
|
||
if idx + 1 < len(data):
|
||
entry_price = float(data[idx + 1]['open'])
|
||
|
||
if signal == "golden": # 做多
|
||
current_position = {
|
||
'direction': 'long',
|
||
'entry_price': entry_price,
|
||
'entry_time': data[idx + 1]['id'],
|
||
'entry_idx': idx + 1,
|
||
'signal': 'golden_cross',
|
||
'highest_price': entry_price, # 用于移动止损
|
||
'lowest_price': entry_price, # 用于空头的移动止损
|
||
}
|
||
self.stats['golden_cross']['count'] += 1
|
||
|
||
elif signal == "dead": # 做空
|
||
current_position = {
|
||
'direction': 'short',
|
||
'entry_price': entry_price,
|
||
'entry_time': data[idx + 1]['id'],
|
||
'entry_idx': idx + 1,
|
||
'signal': 'dead_cross',
|
||
'highest_price': entry_price,
|
||
'lowest_price': entry_price,
|
||
}
|
||
self.stats['dead_cross']['count'] += 1
|
||
|
||
# 跳过下一根，因为已经在这根K线开盘入场
|
||
continue
|
||
|
||
# ========== 持仓时处理 ==========
|
||
if current_position:
|
||
pos_dir = current_position['direction']
|
||
entry_price = current_position['entry_price']
|
||
signal_key = current_position['signal']
|
||
|
||
# 更新最高/最低价（用于移动止损）
|
||
if pos_dir == 'long':
|
||
current_position['highest_price'] = max(current_position['highest_price'], current_price)
|
||
else: # short
|
||
current_position['lowest_price'] = min(current_position['lowest_price'], current_price)
|
||
|
||
# ========== 止损止盈检查 ==========
|
||
should_exit = False
|
||
exit_reason = ""
|
||
exit_price = current_price # 默认用收盘价平仓
|
||
|
||
# 固定止损
|
||
if pos_dir == 'long':
|
||
stop_loss_price = entry_price * (1 - self.stop_loss_pct)
|
||
if current_price <= stop_loss_price:
|
||
should_exit = True
|
||
exit_reason = "止损"
|
||
exit_price = stop_loss_price
|
||
|
||
# 固定止盈
|
||
take_profit_price = entry_price * (1 + self.take_profit_pct)
|
||
if current_price >= take_profit_price:
|
||
should_exit = True
|
||
exit_reason = "止盈"
|
||
exit_price = take_profit_price
|
||
|
||
# 移动止损
|
||
trailing_stop_price = current_position['highest_price'] * (1 - self.trailing_stop_pct)
|
||
if current_price <= trailing_stop_price:
|
||
should_exit = True
|
||
exit_reason = "移动止损"
|
||
exit_price = trailing_stop_price
|
||
|
||
else: # short
|
||
stop_loss_price = entry_price * (1 + self.stop_loss_pct)
|
||
if current_price >= stop_loss_price:
|
||
should_exit = True
|
||
exit_reason = "止损"
|
||
exit_price = stop_loss_price
|
||
|
||
# 固定止盈
|
||
take_profit_price = entry_price * (1 - self.take_profit_pct)
|
||
if current_price <= take_profit_price:
|
||
should_exit = True
|
||
exit_reason = "止盈"
|
||
exit_price = take_profit_price
|
||
|
||
# 移动止损
|
||
trailing_stop_price = current_position['lowest_price'] * (1 + self.trailing_stop_pct)
|
||
if current_price >= trailing_stop_price:
|
||
should_exit = True
|
||
exit_reason = "移动止损"
|
||
exit_price = trailing_stop_price
|
||
|
||
# ========== 反向信号检查 ==========
|
||
if signal and (
|
||
(signal == "dead" and pos_dir == "long") or
|
||
(signal == "golden" and pos_dir == "short")
|
||
):
|
||
should_exit = True
|
||
exit_reason = "反向信号"
|
||
# 反向信号用下一根开盘价平仓
|
||
if idx + 1 < len(data):
|
||
exit_price = float(data[idx + 1]['open'])
|
||
|
||
# ========== 执行平仓 ==========
|
||
if should_exit:
|
||
# 计算盈亏
|
||
if pos_dir == 'long':
|
||
diff = exit_price - entry_price
|
||
else: # short
|
||
diff = entry_price - exit_price
|
||
|
||
# 记录交易
|
||
trade = {
|
||
'entry_time': datetime.datetime.fromtimestamp(current_position['entry_time']),
|
||
'exit_time': datetime.datetime.fromtimestamp(current_bar['id']),
|
||
'signal': self.stats[signal_key]['name'],
|
||
'direction': '做多' if pos_dir == 'long' else '做空',
|
||
'entry': entry_price,
|
||
'exit': exit_price,
|
||
'diff': diff,
|
||
'exit_reason': exit_reason,
|
||
'holding_bars': idx - current_position['entry_idx'] + 1,
|
||
}
|
||
trades.append(trade)
|
||
|
||
# 更新统计
|
||
self.stats[signal_key]['total_profit'] += diff
|
||
if diff > 0:
|
||
self.stats[signal_key]['wins'] += 1
|
||
|
||
# 平仓
|
||
current_position = None
|
||
|
||
# 如果是因为反向信号平仓，立即反手开仓
|
||
if exit_reason == "反向信号" and signal and idx + 1 < len(data):
|
||
if signal == "golden": # 反手做多
|
||
current_position = {
|
||
'direction': 'long',
|
||
'entry_price': exit_price, # 同价反手
|
||
'entry_time': data[idx + 1]['id'],
|
||
'entry_idx': idx + 1,
|
||
'signal': 'golden_cross',
|
||
'highest_price': exit_price,
|
||
'lowest_price': exit_price,
|
||
}
|
||
self.stats['golden_cross']['count'] += 1
|
||
elif signal == "dead": # 反手做空
|
||
current_position = {
|
||
'direction': 'short',
|
||
'entry_price': exit_price,
|
||
'entry_time': data[idx + 1]['id'],
|
||
'entry_idx': idx + 1,
|
||
'signal': 'dead_cross',
|
||
'highest_price': exit_price,
|
||
'lowest_price': exit_price,
|
||
}
|
||
self.stats['dead_cross']['count'] += 1
|
||
|
||
# 跳过下一根K线
|
||
continue
|
||
|
||
# ========== 尾仓处理 ==========
|
||
if current_position:
|
||
last_bar = data[-1]
|
||
exit_price = float(last_bar['close'])
|
||
pos_dir = current_position['direction']
|
||
entry_price = current_position['entry_price']
|
||
signal_key = current_position['signal']
|
||
|
||
diff = (exit_price - entry_price) if pos_dir == 'long' else (entry_price - exit_price)
|
||
|
||
trade = {
|
||
'entry_time': datetime.datetime.fromtimestamp(current_position['entry_time']),
|
||
'exit_time': datetime.datetime.fromtimestamp(last_bar['id']),
|
||
'signal': self.stats[signal_key]['name'],
|
||
'direction': '做多' if pos_dir == 'long' else '做空',
|
||
'entry': entry_price,
|
||
'exit': exit_price,
|
||
'diff': diff,
|
||
'exit_reason': "尾仓平仓",
|
||
'holding_bars': len(data) - current_position['entry_idx'],
|
||
}
|
||
trades.append(trade)
|
||
|
||
self.stats[signal_key]['total_profit'] += diff
|
||
if diff > 0:
|
||
self.stats[signal_key]['wins'] += 1
|
||
|
||
return trades, self.stats
|
||
|
||
|
||
# ========================= 可视化分析 =========================
|
||
|
||
def analyze_trades(trades: List[Dict], stats: Dict) -> Dict:
|
||
"""深入分析交易结果"""
|
||
if not trades:
|
||
return {}
|
||
|
||
# 基础统计
|
||
total_trades = len(trades)
|
||
winning_trades = [t for t in trades if t['diff'] > 0]
|
||
losing_trades = [t for t in trades if t['diff'] <= 0]
|
||
|
||
win_rate = len(winning_trades) / total_trades * 100 if total_trades > 0 else 0
|
||
|
||
# 盈亏统计
|
||
total_profit = sum(t['diff'] for t in trades)
|
||
avg_profit_per_trade = total_profit / total_trades if total_trades > 0 else 0
|
||
|
||
# 胜率相关
|
||
avg_win = np.mean([t['diff'] for t in winning_trades]) if winning_trades else 0
|
||
avg_loss = np.mean([t['diff'] for t in losing_trades]) if losing_trades else 0
|
||
|
||
# 盈亏比
|
||
profit_factor = abs(avg_win / avg_loss) if avg_loss != 0 else float('inf')
|
||
|
||
# 最大连续盈利/亏损
|
||
max_consecutive_wins = 0
|
||
max_consecutive_losses = 0
|
||
current_wins = 0
|
||
current_losses = 0
|
||
|
||
for trade in trades:
|
||
if trade['diff'] > 0:
|
||
current_wins += 1
|
||
current_losses = 0
|
||
max_consecutive_wins = max(max_consecutive_wins, current_wins)
|
||
else:
|
||
current_losses += 1
|
||
current_wins = 0
|
||
max_consecutive_losses = max(max_consecutive_losses, current_losses)
|
||
|
||
# 持仓时间分析
|
||
holding_bars = [t.get('holding_bars', 0) for t in trades]
|
||
avg_holding_bars = np.mean(holding_bars) if holding_bars else 0
|
||
|
||
# 按平仓原因分析
|
||
exit_reasons = {}
|
||
for trade in trades:
|
||
reason = trade.get('exit_reason', '未知')
|
||
exit_reasons[reason] = exit_reasons.get(reason, 0) + 1
|
||
|
||
return {
|
||
'total_trades': total_trades,
|
||
'win_rate': win_rate,
|
||
'total_profit': total_profit,
|
||
'avg_profit_per_trade': avg_profit_per_trade,
|
||
'avg_win': avg_win,
|
||
'avg_loss': avg_loss,
|
||
'profit_factor': profit_factor,
|
||
'max_consecutive_wins': max_consecutive_wins,
|
||
'max_consecutive_losses': max_consecutive_losses,
|
||
'avg_holding_bars': avg_holding_bars,
|
||
'exit_reasons': exit_reasons,
|
||
'winning_trades_count': len(winning_trades),
|
||
'losing_trades_count': len(losing_trades),
|
||
}
|
||
|
||
|
||
# ========================= 主程序 =========================
|
||
|
||
if __name__ == '__main__':
|
||
# 从CSV文件读取数据
|
||
csv_file = "kline_3.csv" # 请替换为你的CSV文件路径
|
||
read_data = []
|
||
|
||
try:
|
||
with open(csv_file, 'r') as csvfile:
|
||
reader = csv.DictReader(csvfile)
|
||
for row in reader:
|
||
read_data.append({
|
||
'id': int(row['id']),
|
||
'open': float(row['open']),
|
||
'high': float(row['high']),
|
||
'low': float(row['low']),
|
||
'close': float(row['close'])
|
||
})
|
||
print(f"成功读取 {len(read_data)} 条K线数据")
|
||
except FileNotFoundError:
|
||
print(f"文件 {csv_file} 未找到，请检查路径")
|
||
exit(1)
|
||
except Exception as e:
|
||
print(f"读取CSV文件时出错: {e}")
|
||
exit(1)
|
||
|
||
# 按时间排序
|
||
read_data.sort(key=lambda x: x['id'])
|
||
|
||
print("\n" + "=" * 60)
|
||
print("EMA交叉策略回测系统")
|
||
print("=" * 60)
|
||
|
||
# 策略参数选择
|
||
print("\n请选择EMA策略类型：")
|
||
print("1. 双EMA交叉策略 (默认: EMA12/EMA26)")
|
||
print("2. 三EMA排列策略 (默认: EMA7/EMA14/EMA30)")
|
||
print("3. EMA+MACD双重确认策略")
|
||
|
||
choice = input("\n请输入选择 (1-3，默认1): ").strip()
|
||
|
||
if choice == "2":
|
||
# 三EMA策略
|
||
backtester = EMABacktester(
|
||
fast_period=7,
|
||
slow_period=30,
|
||
use_triple_ema=True,
|
||
mid_period=14,
|
||
use_macd_confirmation=False,
|
||
stop_loss_pct=0.01, # 1.5%止损
|
||
take_profit_pct=0.4, # 4%止盈
|
||
trailing_stop_pct=0.04 # 2%移动止损
|
||
)
|
||
strategy_name = "三EMA排列策略(7/14/30)"
|
||
|
||
elif choice == "3":
|
||
# EMA+MACD策略
|
||
backtester = EMABacktester(
|
||
fast_period=12,
|
||
slow_period=26,
|
||
signal_period=9,
|
||
use_triple_ema=False,
|
||
use_macd_confirmation=True,
|
||
stop_loss_pct=0.1, # 1%止损
|
||
take_profit_pct=0.4, # 3%止盈
|
||
trailing_stop_pct=0.04 # 1.5%移动止损
|
||
)
|
||
strategy_name = "EMA+MACD双重确认策略(12/26/9)"
|
||
|
||
else:
|
||
# 默认双EMA策略
|
||
backtester = EMABacktester(
|
||
fast_period=12,
|
||
slow_period=26,
|
||
use_triple_ema=False,
|
||
use_macd_confirmation=False,
|
||
stop_loss_pct=0.02, # 2%止损
|
||
take_profit_pct=0.05, # 5%止盈
|
||
trailing_stop_pct=0.03 # 3%移动止损
|
||
)
|
||
strategy_name = "双EMA交叉策略(12/26)"
|
||
|
||
print(f"\n使用策略: {strategy_name}")
|
||
print("开始回测...")
|
||
|
||
# 执行回测
|
||
trades, stats = backtester.backtest(read_data)
|
||
|
||
# ========== 交易详情和盈利计算 ==========
|
||
print(f"\n{'=' * 60}")
|
||
print(f"回测结果 - {strategy_name}")
|
||
print(f"{'=' * 60}")
|
||
|
||
# 参数设定
|
||
contract_size = 10000 # 合约规模
|
||
open_fee_fixed = 5 # 固定开仓手续费
|
||
close_fee_rate = 0.0005 # 平仓手续费率
|
||
|
||
total_points_profit = 0 # 累计点差
|
||
total_money_profit = 0 # 累计金额盈利
|
||
total_fee = 0 # 累计手续费
|
||
|
||
print(f"\n交易详情 (共{len(trades)}笔):")
|
||
print("-" * 100)
|
||
|
||
for i, t in enumerate(trades, 1):
|
||
entry = t['entry']
|
||
exit = t['exit']
|
||
direction = t['direction']
|
||
|
||
# 原始价差
|
||
point_diff = (exit - entry) if direction == '做多' else (entry - exit)
|
||
|
||
# 金额盈利
|
||
money_profit = point_diff / entry * contract_size
|
||
|
||
# 手续费
|
||
fee = open_fee_fixed + (contract_size / entry * exit * close_fee_rate)
|
||
|
||
# 净利润
|
||
net_profit = money_profit - fee
|
||
|
||
# 保存结果
|
||
t.update({
|
||
'point_diff': point_diff,
|
||
'raw_profit': money_profit,
|
||
'fee': fee,
|
||
'net_profit': net_profit
|
||
})
|
||
|
||
total_points_profit += point_diff
|
||
total_money_profit += money_profit
|
||
total_fee += fee
|
||
|
||
# 输出交易详情
|
||
profit_color = "\033[92m" if net_profit > 0 else "\033[91m"
|
||
reset_color = "\033[0m"
|
||
|
||
print(f"{i:3d}. {t['entry_time'].strftime('%Y-%m-%d %H:%M')} -> "
|
||
f"{t['exit_time'].strftime('%Y-%m-%d %H:%M')} "
|
||
f"{direction}({t['signal']}) "
|
||
f"入={entry:.2f} 出={exit:.2f} "
|
||
f"{profit_color}净利={net_profit:+.2f}{reset_color} "
|
||
f"(持有:{t.get('holding_bars', '?')}根K线, 原因:{t.get('exit_reason', '未知')})")
|
||
|
||
# ========== 汇总统计 ==========
|
||
total_net_profit = total_money_profit - total_fee
|
||
|
||
print(f"\n{'=' * 60}")
|
||
print("汇总统计")
|
||
print(f"{'=' * 60}")
|
||
|
||
print(f"总交易笔数: {len(trades)}")
|
||
print(f"总点差: {total_points_profit:.2f}")
|
||
print(f"总原始盈利（未扣费）: {total_money_profit:.2f}")
|
||
print(f"总手续费: {total_fee:.2f}")
|
||
print(f"总净利润: {total_net_profit:.2f}")
|
||
|
||
# 深入分析
|
||
analysis = analyze_trades(trades, stats)
|
||
|
||
if analysis:
|
||
print(f"\n策略分析:")
|
||
print(f"- 胜率: {analysis['win_rate']:.2f}%")
|
||
print(f"- 平均每笔盈利: {analysis['avg_profit_per_trade']:.2f}")
|
||
print(f"- 平均盈利: {analysis['avg_win']:.2f}")
|
||
print(f"- 平均亏损: {analysis['avg_loss']:.2f}")
|
||
print(f"- 盈亏比: {analysis['profit_factor']:.2f}")
|
||
print(f"- 最大连续盈利: {analysis['max_consecutive_wins']} 笔")
|
||
print(f"- 最大连续亏损: {analysis['max_consecutive_losses']} 笔")
|
||
print(f"- 平均持仓K线数: {analysis['avg_holding_bars']:.1f} 根")
|
||
|
||
print(f"\n平仓原因统计:")
|
||
for reason, count in analysis['exit_reasons'].items():
|
||
percentage = count / len(trades) * 100
|
||
print(f" - {reason}: {count} 笔 ({percentage:.1f}%)")
|
||
|
||
# ========== 信号统计 ==========
|
||
print(f"\n{'=' * 60}")
|
||
print("信号统计")
|
||
print(f"{'=' * 60}")
|
||
|
||
for k, v in stats.items():
|
||
name, count, wins, total_p = v['name'], v['count'], v['wins'], v['total_profit']
|
||
if count > 0:
|
||
win_rate = (wins / count * 100)
|
||
avg_p = total_p / count
|
||
profit_color = "\033[92m" if total_p > 0 else "\033[91m"
|
||
reset_color = "\033[0m"
|
||
|
||
print(f"{name}:")
|
||
print(f" 信号次数: {count}")
|
||
print(f" 胜率: {win_rate:.2f}%")
|
||
print(f" 总价差: {profit_color}{total_p:.2f}{reset_color}")
|
||
print(f" 平均价差: {avg_p:.2f}")
|
||
|
||
# ========== 风险指标 ==========
|
||
if len(trades) > 1:
|
||
returns = [t['net_profit'] for t in trades]
|
||
|
||
# 夏普比率（简化版）
|
||
avg_return = np.mean(returns)
|
||
std_return = np.std(returns)
|
||
sharpe_ratio = avg_return / std_return if std_return > 0 else 0
|
||
|
||
# 最大回撤
|
||
cumulative_returns = np.cumsum(returns)
|
||
running_max = np.maximum.accumulate(cumulative_returns)
|
||
drawdown = cumulative_returns - running_max
|
||
max_drawdown = abs(np.min(drawdown)) if len(drawdown) > 0 else 0
|
||
|
||
print(f"\n{'=' * 60}")
|
||
print("风险指标")
|
||
print(f"{'=' * 60}")
|
||
print(f"夏普比率（简化）: {sharpe_ratio:.4f}")
|
||
print(f"最大回撤: {max_drawdown:.2f}")
|
||
|
||
# 盈亏分布
|
||
print(f"\n盈亏分布:")
|
||
profit_ranges = {
|
||
"大亏 (< -100)": len([r for r in returns if r < -100]),
|
||
"中亏 (-100 ~ -50)": len([r for r in returns if -100 <= r < -50]),
|
||
"小亏 (-50 ~ 0)": len([r for r in returns if -50 <= r < 0]),
|
||
"小盈 (0 ~ 50)": len([r for r in returns if 0 <= r < 50]),
|
||
"中盈 (50 ~ 100)": len([r for r in returns if 50 <= r < 100]),
|
||
"大盈 (> 100)": len([r for r in returns if r >= 100]),
|
||
}
|
||
|
||
for range_name, count in profit_ranges.items():
|
||
if count > 0:
|
||
percentage = count / len(returns) * 100
|
||
print(f" {range_name}: {count} 笔 ({percentage:.1f}%)")
|
||
|
||
print(f"\n{'=' * 60}")
|
||
print("回测完成!")
|
||
print(f"{'=' * 60}") |