commit code

策略1：技术指标MACD 金叉+MA 多头.py

+instruments =['000021.SZA','000034.SZA','000066.SZA','000158.SZA','000555.SZA','000606.SZA','000662.SZA','000938.SZA','000948.SZA','000606.SZA','300157.SZA','300164.SZA','300191.SZA','600121.SHA','600123.SHA','600157.SHA','600188.SHA','600193.SHA','600295.SHA','600339.SHA','300610.SZA','300637.SZA','300641.SZA','300655.SZA','300665.SZA','300690.SZA','300699.SZA','300716.SZA','300717.SZA','300721.SZA','000060.SZA','000426.SZA','000511.SZA','000603.SZA','000612.SZA','000630.SZA','000633.SZA','000657.SZA','000688.SZA','000693.SZA','300592.SZA','300621.SZA','300635.SZA','300649.SZA','300668.SZA','600039.SHA','600068.SHA','600133.SHA','600170.SHA','600209.SHA','603218.SHA','603320.SHA','603333.SHA','603396.SHA','603416.SHA','603488.SHA','603507.SHA','603577.SHA','603606.SHA','603488.SHA','300403.SZA','300475.SZA','600060.SHA','600336.SHA','600619.SHA','600690.SHA','600839.SHA','600854.SHA','600870.SHA','600983.SHA','600867.SHA','600896.SHA','600976.SHA','600993.SHA','600998.SHA','601607.SHA','603079.SHA','603108.SHA','603127.SHA','603139.SHA','002351.SZA','002369.SZA','002371.SZA','002388.SZA','002384.SZA','002402.SZA','002414.SZA','002463.SZA','002475.SZA','002484.SZA','002711.SZA','002769.SZA','002800.SZA','002889.SZA','300013.SZA','300240.SZA','300350.SZA','600004.SHA','600009.SHA','600026.SHA','601128.SHA','601166.SHA','601169.SHA','601229.SHA','601288.SHA','601398.SHA','601328.SHA','601818.SHA','601939.SHA','601988.SHA','600804.SHA','603042.SHA','603083.SHA','603118.SHA','603322.SHA','603421.SHA','603559.SHA','603602.SHA','603703.SHA','603803.SHA']
+instruments =['000021.SZA','000034.SZA','000066.SZA','000158.SZA','000555.SZA','000606.SZA','000662.SZA','000938.SZA','000948.SZA','000606.SZA','300157.SZA','300164.SZA','300191.SZA','600121.SHA','600123.SHA','600157.SHA','600188.SHA','600193.SHA','600295.SHA','600339.SHA','300610.SZA','300637.SZA','300641.SZA','300655.SZA','300665.SZA','300690.SZA','300699.SZA','300716.SZA','300717.SZA','300721.SZA','000060.SZA','000426.SZA','000511.SZA','000603.SZA','000612.SZA','000630.SZA','000633.SZA','000657.SZA','000688.SZA','000693.SZA','300592.SZA','300621.SZA','300635.SZA','300649.SZA','300668.SZA','600039.SHA','600068.SHA','600133.SHA','600170.SHA','600209.SHA','603218.SHA','603320.SHA','603333.SHA','603396.SHA','603416.SHA','603488.SHA','603507.SHA','603577.SHA','603606.SHA','603488.SHA','300403.SZA','300475.SZA','600060.SHA','600336.SHA','600619.SHA','600690.SHA','600839.SHA','600854.SHA','600870.SHA','600983.SHA','600867.SHA','600896.SHA','600976.SHA','600993.SHA','600998.SHA','601607.SHA','603079.SHA','603108.SHA','603127.SHA','603139.SHA','002351.SZA','002369.SZA','002371.SZA','002388.SZA','002384.SZA','002402.SZA','002414.SZA','002463.SZA','002475.SZA','002484.SZA','002711.SZA','002769.SZA','002800.SZA','002889.SZA','300013.SZA','300240.SZA','300350.SZA','600004.SHA','600009.SHA','600026.SHA','601128.SHA','601166.SHA','601169.SHA','601229.SHA','601288.SHA','601398.SHA','601328.SHA','601818.SHA','601939.SHA','601988.SHA','600804.SHA','603042.SHA','603083.SHA','603118.SHA','603322.SHA','603421.SHA','603559.SHA','603602.SHA','603703.SHA','603803.SHA']
+
+start_date = '20190103'
+end_date = '20210122'
+
+import json
+import talib
+import numpy as np
+
+
+def initialize(context):
+    # MACD 金叉MA 多头 指标设置DIF 短线=12, DIF 长线=26, DEA=9
+    context.short_period = 12
+    context.long_period = 26
+    context.dea = 9
+
+    context.trading_day_count = 0
+
+    # MA 多头 MA 短线=5，MA 长线=20
+    context.ma_short = 5
+    context.ma_long = 20
+
+    # 资金
+    context.cash = 10000
+
+    # 交易费用
+    context.set_commission(PerOrder(buy_cost=0.0003, sell_cost=0.0003, min_cost=5))
+
+
+def is_all_nan(data):
+    for datum in data:
+        if str(datum) != 'nan':
+            return False
+    return True
+
+
+# 计算MACD
+def macd(context, data):
+    stock_set = set()
+    d = data.history(data.keys(), 'price', bar_count=34, frequency='1d')
+    for key in data.keys():
+        s = np.asarray(d[key])
+        if not is_all_nan(s):
+            macd, macdsignal, macdhist = talib.MACD(s, context.short_period, context.long_period, context.dea)
+            if macdhist[-1] > 0:
+                stock_set.add(key.symbol)
+    return stock_set
+
+
+# 计算MA
+def ma(context, data):
+    stock_set = set()
+    d_5 = data.history(data.keys(), 'price', bar_count=5, frequency='1d')
+    d_20 = data.history(data.keys(), 'price', bar_count=20, frequency='1d')
+
+    for key in data.keys():
+        s_5 = np.asarray(d_5[key])
+        s_20 = np.asarray(d_20[key])
+        if not is_all_nan(s_5) and not is_all_nan(s_20):
+            ma5 = talib.MA(s_5, timeperiod=5, matype=0)
+            ma20 = talib.MA(s_20, timeperiod=20, matype=0)
+            if ma5[-1] > ma20[-1]:
+                stock_set.add(key.symbol)
+
+    return stock_set
+
+
+def handle_data(context, data):
+    # 每20日交易一次
+    if context.trading_day_count % 20 == 0:
+
+        set_macd = macd(context, data)
+        set_ma = ma(context, data)
+        candidate_list = list(set_macd & set_ma)
+        max_stock_count = 10
+        if len(candidate_list) > max_stock_count:
+            candidate_list = candidate_list[0:10]
+
+        # 全部卖出
+        for equity in context.portfolio.positions:
+            position = context.portfolio.positions[equity]
+            sid = position.sid
+            if data.can_trade(sid):
+                context.order_target_percent(sid, 0)
+
+        cash = context.portfolio.cash
+
+        # 买入
+        for stock in candidate_list:
+            sid = context.symbol(stock)
+            if data.can_trade(sid):
+                price = data.current(sid, 'price')
+                context.order(sid, int(cash / price / 100 / len(candidate_list)) * 100)
+
+                # 交易日+1
+        context.trading_day_count = context.trading_day_count + 1
+
+
+m=M.trade.v2(
+    instruments=instruments,
+    start_date=start_date,
+    end_date=end_date,
+    initialize=initialize,
+    handle_data=handle_data,
+    order_price_field_buy='open', # 以开盘价买入
+    order_price_field_sell='open', # 以开盘价卖出
+    capital_base=100000, # 本金
+    m_deps=np.random.randn()
+    )
\ No newline at end of file

策略2：技术指标自定义策略.py

+instruments =['000021.SZA','000034.SZA','000066.SZA','000158.SZA','000555.SZA','000606.SZA','000662.SZA','000938.SZA','000948.SZA','000606.SZA','300157.SZA','300164.SZA','300191.SZA','600121.SHA','600123.SHA','600157.SHA','600188.SHA','600193.SHA','600295.SHA','600339.SHA','300610.SZA','300637.SZA','300641.SZA','300655.SZA','300665.SZA','300690.SZA','300699.SZA','300716.SZA','300717.SZA','300721.SZA','000060.SZA','000426.SZA','000511.SZA','000603.SZA','000612.SZA','000630.SZA','000633.SZA','000657.SZA','000688.SZA','000693.SZA','300592.SZA','300621.SZA','300635.SZA','300649.SZA','300668.SZA','600039.SHA','600068.SHA','600133.SHA','600170.SHA','600209.SHA','603218.SHA','603320.SHA','603333.SHA','603396.SHA','603416.SHA','603488.SHA','603507.SHA','603577.SHA','603606.SHA','603488.SHA','300403.SZA','300475.SZA','600060.SHA','600336.SHA','600619.SHA','600690.SHA','600839.SHA','600854.SHA','600870.SHA','600983.SHA','600867.SHA','600896.SHA','600976.SHA','600993.SHA','600998.SHA','601607.SHA','603079.SHA','603108.SHA','603127.SHA','603139.SHA','002351.SZA','002369.SZA','002371.SZA','002388.SZA','002384.SZA','002402.SZA','002414.SZA','002463.SZA','002475.SZA','002484.SZA','002711.SZA','002769.SZA','002800.SZA','002889.SZA','300013.SZA','300240.SZA','300350.SZA','600004.SHA','600009.SHA','600026.SHA','601128.SHA','601166.SHA','601169.SHA','601229.SHA','601288.SHA','601398.SHA','601328.SHA','601818.SHA','601939.SHA','601988.SHA','600804.SHA','603042.SHA','603083.SHA','603118.SHA','603322.SHA','603421.SHA','603559.SHA','603602.SHA','603703.SHA','603803.SHA']
+instruments =['000021.SZA','000034.SZA','000066.SZA','000158.SZA','000555.SZA','000606.SZA','000662.SZA','000938.SZA','000948.SZA','000606.SZA','300157.SZA','300164.SZA','300191.SZA','600121.SHA','600123.SHA','600157.SHA','600188.SHA','600193.SHA','600295.SHA','600339.SHA','300610.SZA','300637.SZA','300641.SZA','300655.SZA','300665.SZA','300690.SZA','300699.SZA','300716.SZA','300717.SZA','300721.SZA','000060.SZA','000426.SZA','000511.SZA','000603.SZA','000612.SZA','000630.SZA','000633.SZA','000657.SZA','000688.SZA','000693.SZA','300592.SZA','300621.SZA','300635.SZA','300649.SZA','300668.SZA','600039.SHA','600068.SHA','600133.SHA','600170.SHA','600209.SHA','603218.SHA','603320.SHA','603333.SHA','603396.SHA','603416.SHA','603488.SHA','603507.SHA','603577.SHA','603606.SHA','603488.SHA','300403.SZA','300475.SZA','600060.SHA','600336.SHA','600619.SHA','600690.SHA','600839.SHA','600854.SHA','600870.SHA','600983.SHA','600867.SHA','600896.SHA','600976.SHA','600993.SHA','600998.SHA','601607.SHA','603079.SHA','603108.SHA','603127.SHA','603139.SHA','002351.SZA','002369.SZA','002371.SZA','002388.SZA','002384.SZA','002402.SZA','002414.SZA','002463.SZA','002475.SZA','002484.SZA','002711.SZA','002769.SZA','002800.SZA','002889.SZA','300013.SZA','300240.SZA','300350.SZA','600004.SHA','600009.SHA','600026.SHA','601128.SHA','601166.SHA','601169.SHA','601229.SHA','601288.SHA','601398.SHA','601328.SHA','601818.SHA','601939.SHA','601988.SHA','600804.SHA','603042.SHA','603083.SHA','603118.SHA','603322.SHA','603421.SHA','603559.SHA','603602.SHA','603703.SHA','603803.SHA']
+
+start_date = '20190103'
+end_date = '20210122'
+
+import json
+import talib
+import numpy as np
+
+
+def initialize(context):
+    # MACD 金叉MA 多头 指标设置DIF 短线=12, DIF 长线=26, DEA=9
+    context.short_period = 12
+    context.long_period = 26
+    context.dea = 9
+
+    context.trading_day_count = 0
+
+    # MA 多头 MA 短线=5，MA 长线=20
+    context.ma_short = 5
+    context.ma_long = 20
+
+    # 资金
+    context.cash = 10000
+
+    # 交易费用
+    context.set_commission(PerOrder(buy_cost=0.0003, sell_cost=0.0003, min_cost=5))
+
+
+def is_all_nan(data):
+    for datum in data:
+        if str(datum) != 'nan':
+            return False
+    return True
+
+
+def min_cap(context, data):
+    market_cap_data = D.history_data(instruments, data.current_dt, data.current_dt,
+                                     fields=['market_cap', 'amount', 'suspended'])
+    # 根据是否停牌的字段确定每日选出来的股票
+    daily_buy_stock = market_cap_data.groupby('date').apply(lambda df: df[(df['amount'] > 0)  # 需要有成交量
+                                                                          & (df['suspended'] == False)  # 是否停牌
+                                                                          ].sort_values('market_cap')[:10])  #
+    # context.logger.info(daily_buy_stock['instrument'])
+    return set(daily_buy_stock['instrument'])
+
+
+# 计算MACD
+def macd(context, data):
+    stock_set = set()
+    d = data.history(data.keys(), 'price', bar_count=34, frequency='1d')
+    for key in data.keys():
+        s = np.asarray(d[key])
+        if not is_all_nan(s):
+            macd, macdsignal, macdhist = talib.MACD(s, context.short_period, context.long_period, context.dea)
+            if macdhist[-1] > 0:
+                stock_set.add(key.symbol)
+    return stock_set
+
+
+# 计算MA
+def ma(context, data):
+    stock_set = set()
+    d_5 = data.history(data.keys(), 'price', bar_count=5, frequency='1d')
+    d_20 = data.history(data.keys(), 'price', bar_count=20, frequency='1d')
+
+    for key in data.keys():
+        s_5 = np.asarray(d_5[key])
+        s_20 = np.asarray(d_20[key])
+        if not is_all_nan(s_5) and not is_all_nan(s_20):
+            ma5 = talib.MA(s_5, timeperiod=5, matype=0)
+            ma20 = talib.MA(s_20, timeperiod=20, matype=0)
+            if ma5[-1] > ma20[-1]:
+                stock_set.add(key.symbol)
+
+    return stock_set
+
+
+def rebalance(context, data):
+    set_cap = min_cap(context, data)
+    # context.logger.info(set_cap)
+    set_macd = macd(context, data)
+    set_ma = ma(context, data)
+    candidate_list = list(set_cap & set_macd & set_ma)
+    max_stock_count = 10
+    if len(candidate_list) > max_stock_count:
+        candidate_list = candidate_list[0:10]
+
+    if len(candidate_list) > 0:
+
+        # 当前持有的股票
+        stock_hold_now = [equity.symbol for equity in context.portfolio.positions]
+
+        # 无需卖出
+        stock_not_to_sell = [i for i in stock_hold_now if i in candidate_list]
+
+        # 需要卖出的股票
+        stock_to_sell = [i for i in stock_hold_now if i not in stock_not_to_sell]
+
+        # 卖出
+        for stock in stock_to_sell:
+            if data.can_trade(context.symbol(stock)):
+                context.order_target_percent(context.symbol(stock), 0)
+
+        # 等权重买入
+        weight = 1 / len(candidate_list)
+
+        # 买入
+        for stock in candidate_list:
+            if data.can_trade(context.symbol(stock)):
+                context.order_target_percent(context.symbol(stock), weight)
+
+    # 交易日+1
+    context.trading_day_count = context.trading_day_count + 1
+
+
+def handle_data(context, data):
+    # 每10日交易一次
+    if context.trading_day_count % 20 == 0:
+        rebalance(context, data)
+
+
+
+
+m=M.trade.v2(
+    instruments=instruments,
+    start_date=start_date,
+    end_date=end_date,
+    initialize=initialize,
+    handle_data=handle_data,
+    order_price_field_buy='open', # 以开盘价买入
+    order_price_field_sell='open', # 以开盘价卖出
+    capital_base=100000, # 本金
+    m_deps=np.random.randn()
+    )
\ No newline at end of file

策略3：QP.py

+instruments =['000021.SZA','000034.SZA','000066.SZA','000158.SZA','000555.SZA','000606.SZA','000662.SZA','000938.SZA','000948.SZA','000606.SZA','300157.SZA','300164.SZA','300191.SZA','600121.SHA','600123.SHA','600157.SHA','600188.SHA','600193.SHA','600295.SHA','600339.SHA','300610.SZA','300637.SZA','300641.SZA','300655.SZA','300665.SZA','300690.SZA','300699.SZA','300716.SZA','300717.SZA','300721.SZA','000060.SZA','000426.SZA','000511.SZA','000603.SZA','000612.SZA','000630.SZA','000633.SZA','000657.SZA','000688.SZA','000693.SZA','300592.SZA','300621.SZA','300635.SZA','300649.SZA','300668.SZA','600039.SHA','600068.SHA','600133.SHA','600170.SHA','600209.SHA','603218.SHA','603320.SHA','603333.SHA','603396.SHA','603416.SHA','603488.SHA','603507.SHA','603577.SHA','603606.SHA','603488.SHA','300403.SZA','300475.SZA','600060.SHA','600336.SHA','600619.SHA','600690.SHA','600839.SHA','600854.SHA','600870.SHA','600983.SHA','600867.SHA','600896.SHA','600976.SHA','600993.SHA','600998.SHA','601607.SHA','603079.SHA','603108.SHA','603127.SHA','603139.SHA','002351.SZA','002369.SZA','002371.SZA','002388.SZA','002384.SZA','002402.SZA','002414.SZA','002463.SZA','002475.SZA','002484.SZA','002711.SZA','002769.SZA','002800.SZA','002889.SZA','300013.SZA','300240.SZA','300350.SZA','600004.SHA','600009.SHA','600026.SHA','601128.SHA','601166.SHA','601169.SHA','601229.SHA','601288.SHA','601398.SHA','601328.SHA','601818.SHA','601939.SHA','601988.SHA','600804.SHA','603042.SHA','603083.SHA','603118.SHA','603322.SHA','603421.SHA','603559.SHA','603602.SHA','603703.SHA','603803.SHA']
+instruments =['000021.SZA','000034.SZA','000066.SZA','000158.SZA','000555.SZA','000606.SZA','000662.SZA','000938.SZA','000948.SZA','000606.SZA','300157.SZA','300164.SZA','300191.SZA','600121.SHA','600123.SHA','600157.SHA','600188.SHA','600193.SHA','600295.SHA','600339.SHA','300610.SZA','300637.SZA','300641.SZA','300655.SZA','300665.SZA','300690.SZA','300699.SZA','300716.SZA','300717.SZA','300721.SZA','000060.SZA','000426.SZA','000511.SZA','000603.SZA','000612.SZA','000630.SZA','000633.SZA','000657.SZA','000688.SZA','000693.SZA','300592.SZA','300621.SZA','300635.SZA','300649.SZA','300668.SZA','600039.SHA','600068.SHA','600133.SHA','600170.SHA','600209.SHA','603218.SHA','603320.SHA','603333.SHA','603396.SHA','603416.SHA','603488.SHA','603507.SHA','603577.SHA','603606.SHA','603488.SHA','300403.SZA','300475.SZA','600060.SHA','600336.SHA','600619.SHA','600690.SHA','600839.SHA','600854.SHA','600870.SHA','600983.SHA','600867.SHA','600896.SHA','600976.SHA','600993.SHA','600998.SHA','601607.SHA','603079.SHA','603108.SHA','603127.SHA','603139.SHA','002351.SZA','002369.SZA','002371.SZA','002388.SZA','002384.SZA','002402.SZA','002414.SZA','002463.SZA','002475.SZA','002484.SZA','002711.SZA','002769.SZA','002800.SZA','002889.SZA','300013.SZA','300240.SZA','300350.SZA','600004.SHA','600009.SHA','600026.SHA','601128.SHA','601166.SHA','601169.SHA','601229.SHA','601288.SHA','601398.SHA','601328.SHA','601818.SHA','601939.SHA','601988.SHA','600804.SHA','603042.SHA','603083.SHA','603118.SHA','603322.SHA','603421.SHA','603559.SHA','603602.SHA','603703.SHA','603803.SHA']
+
+start_date = '20190103'
+end_date = '20210122'
+
+import json
+import talib
+import numpy as np
+import cvxopt as opt
+from cvxopt import blas, solvers
+
+solvers.options['show_progress'] = False
+
+def initialize(context):
+
+    context.set_commission(PerOrder(buy_cost=0.003, sell_cost=0.003, min_cost=5))
+
+
+def optimal_portfolio(returns):
+    returns = np.asmatrix(returns)
+    n = len(returns)
+
+    # 转化为cvxopt matrices
+    # Σ: 这是收益的协方差矩阵，代表着股票之间的关系
+    S = opt.matrix(np.cov(returns))
+    # μ = (μ1, ..., μn)：这是通过历史数据计算出来的每只股票的期望收益。
+    pbar = opt.matrix(np.mean(returns, axis=1))
+
+    # 约束条件
+    G = -opt.matrix(np.eye(n))
+    h = opt.matrix(0.0, (n, 1))
+    A = opt.matrix(1.0, (1, n))
+    b = opt.matrix(1.0)
+
+    # 使用凸优化计算有效前沿
+    # 计算最优组合
+    wt = solvers.qp(opt.matrix(0.1 * S), -pbar, G, h, A, b)['x']
+    return np.asarray(wt)
+
+
+def handle_data(context, data):
+    if context.trading_day_index % 20 == 0:
+
+        prices = data.history(data.keys(), fields='price', bar_count=6, frequency='1d').dropna(
+            axis=1).pct_change().dropna(axis=0)
+
+        if not prices.empty:
+            weights = optimal_portfolio(prices.T)
+            for stock, weight in zip(prices.columns, weights):
+                if data.can_trade(stock):
+                    context.order_target_percent(stock, weight[0])
+
+m=M.trade.v2(
+    instruments=instruments,
+    start_date=start_date,
+    end_date=end_date,
+    initialize=initialize,
+    handle_data=handle_data,
+    order_price_field_buy='open', # 以开盘价买入
+    order_price_field_sell='open', # 以开盘价卖出
+    capital_base=100000, # 本金
+    m_deps=np.random.randn()
+    )
\ No newline at end of file

策略4：QP + 多样性.py

+
+
+
+
+start_date = '20190103'
+end_date = '20210122'
+
+import pandas as pd
+import json
+import talib
+import numpy as np
+import cvxopt as opt
+from cvxopt import blas, solvers
+
+solvers.options['show_progress'] = False
+
+# 股票
+df = pd.read_csv('stock_list.csv',header=None)
+instruments = df[0].to_list()
+
+# 每个行业下股票数量
+industries = df[1].to_list()
+industries_cnt = [industries.count(i) for i in np.unique(industries)]
+
+
+def initialize(context):
+    # 设置手续费
+    context.set_commission(PerOrder(buy_cost=0.003, sell_cost=0.003, min_cost=5))
+
+    context.rebalance_period = 20
+    context.observation = 5
+    context.max_weight_per_stock = 0.1
+    context.max_weight_per_industry = 0.3
+
+
+def optimal_portfolio(context, returns):
+    returns = np.asmatrix(returns)
+    n = len(returns)
+
+    # 转化为cvxopt matrices
+    S = opt.matrix(np.cov(returns))
+    pbar = opt.matrix(np.mean(returns, axis=1))
+
+    # 初始约束条件
+    G1 = -opt.matrix(np.eye(n))
+    h1 = opt.matrix(0.0, (n, 1))
+    # 每只股票权重限制
+    G2 = opt.matrix(np.eye(n))
+    h2 = opt.matrix(context.max_weight_per_stock, (n, 1))
+    # 行业限制
+    G3 = np.zeros((len(industries_cnt), n))
+    start = 0
+    end = 0
+    for i in range(len(industries_cnt)):
+        end += industries_cnt[i]
+        G3[i, start:end] = 1
+        start = end
+    G3 = opt.matrix(G3)
+    h3 = opt.matrix(context.max_weight_per_industry, (len(industries_cnt), 1))
+    G = opt.matrix(np.concatenate([G1, G2, G3]))
+    h = opt.matrix(np.concatenate([h1, h2, h3]))
+    A = opt.matrix(1.0, (1, n))
+    b = opt.matrix(1.0)
+
+    # 使用凸优化计算有效前沿
+    # 计算最优组合
+    wt = solvers.qp(opt.matrix(0.1 * S), -pbar, G, h, A, b)['x']
+    return np.asarray(wt)
+
+
+def handle_data(context, data):
+    if context.trading_day_index % 20 == 0:
+
+        prices = data.history(data.keys(), fields='price', bar_count=6, frequency='1d').dropna(
+            axis=1).pct_change().dropna(axis=0)
+
+        if not prices.empty:
+            weights = optimal_portfolio(context, prices.T)
+            for stock, weight in zip(prices.columns, weights):
+                if data.can_trade(stock):
+                    context.order_target_percent(stock, weight[0])
+
+m=M.trade.v2(
+    instruments=instruments,
+    start_date=start_date,
+    end_date=end_date,
+    initialize=initialize,
+    handle_data=handle_data,
+    order_price_field_buy='open', # 以开盘价买入
+    order_price_field_sell='open', # 以开盘价卖出
+    capital_base=100000, # 本金
+    m_deps=np.random.randn()
+    )
\ No newline at end of file

策略7：QP + 自定义.py

+
+
+start_date = '20190103'
+end_date = '20210122'
+
+import pandas as pd
+import json
+import talib
+import numpy as np
+import cvxopt as opt
+from cvxopt import blas, solvers
+
+solvers.options['show_progress'] = False
+
+# 股票
+df = pd.read_csv('stock_list.csv',header=None)
+instruments = df[0].to_list()
+
+# 每个行业下股票数量
+industries = df[1].to_list()
+industries_cnt = [industries.count(i) for i in np.unique(industries)]
+
+
+def initialize(context):
+    # 设置手续费
+    context.set_commission(PerOrder(buy_cost=0.003, sell_cost=0.003, min_cost=5))
+
+    context.rebalance_period = 20
+    context.observation = 5
+    context.max_weight_per_stock = 0.1
+    context.max_weight_per_industry = 0.3
+
+
+def optimal_portfolio(context, returns):
+    returns = np.asmatrix(returns)
+    n = len(returns)
+
+    # 转化为cvxopt matrices
+    S = opt.matrix(np.cov(returns))
+    pbar = opt.matrix(np.mean(returns, axis=1))
+
+    # 初始约束条件
+    G1 = -opt.matrix(np.eye(n))  # opt默认是求最大值，因此要求最小化问题，还得乘以一个负号
+    h1 = opt.matrix(0.0, (n, 1))
+    # 每只股票权重限制
+    G2 = opt.matrix(np.eye(n))
+    h2 = opt.matrix(context.max_weight_per_stock, (n, 1))
+    # 行业限制
+    G3 = np.zeros((len(industries_cnt), n))
+    start = 0
+    end = 0
+    for i in range(len(industries_cnt)):
+        end += industries_cnt[i]
+        G3[i, start:end] = 1
+        start = end
+    G3 = opt.matrix(G3)
+    h3 = opt.matrix(context.max_weight_per_industry, (len(industries_cnt), 1))
+    G = opt.matrix(np.concatenate([G1, G2, G3]))
+    h = opt.matrix(np.concatenate([h1, h2, h3]))
+    A = opt.matrix(1.0, (1, n))
+    b = opt.matrix(1.0)
+
+    # 使用凸优化计算有效前沿
+    # 计算最优组合
+    wt = solvers.qp(opt.matrix(0.1 * S), -pbar, G, h, A, b)['x']
+    return np.asarray(wt)
+
+
+def is_all_nan(data):
+    for datum in data:
+        if str(datum) != 'nan':
+            return False
+    return True
+
+
+def min_cap(context, data):
+    market_cap_data = D.history_data(instruments, data.current_dt, data.current_dt,
+                                     fields=['market_cap', 'amount', 'suspended'])
+    # 根据是否停牌的字段确定每日选出来的股票
+    daily_buy_stock = market_cap_data.groupby('date').apply(lambda df: df[(df['amount'] > 0)  # 需要有成交量
+                                                                          & (df['suspended'] == False)  # 是否停牌
+                                                                          ].sort_values('market_cap')[:30])  #
+    # context.logger.info(daily_buy_stock['instrument'])
+    return set(daily_buy_stock['instrument'])
+
+
+# 计算MA
+def ma(context, data):
+    stock_set = set()
+    d_5 = data.history(data.keys(), 'price', bar_count=5, frequency='1d')
+    d_20 = data.history(data.keys(), 'price', bar_count=20, frequency='1d')
+
+    for key in data.keys():
+        s_5 = np.asarray(d_5[key])
+        s_20 = np.asarray(d_20[key])
+        if not is_all_nan(s_5) and not is_all_nan(s_20):
+            ma5 = talib.MA(s_5, timeperiod=5, matype=0)
+            ma20 = talib.MA(s_20, timeperiod=20, matype=0)
+            if ma5[-1] > ma20[-1]:
+                stock_set.add(key.symbol)
+
+    return stock_set
+
+
+def handle_data(context, data):
+    if context.trading_day_index % 20 == 0:
+
+        set_cap = min_cap(context, data)
+        set_ma = ma(context, data)
+
+        candidate_list = list(set_cap & set_ma)
+
+        candidate = []
+
+        for stock in candidate_list:
+            candidate.append(context.symbol(stock))
+
+        if len(candidate) > 1:
+            prices = data.history(candidate, fields='price', bar_count=6, frequency='1d').dropna(
+                axis=1).pct_change().dropna(axis=0)
+
+            if not prices.empty:
+                weights = optimal_portfolio(context, prices.T)
+                for stock, weight in zip(prices.columns, weights):
+                    if data.can_trade(stock):
+                        context.order_target_percent(stock, weight[0])
+
+m=M.trade.v2(
+    instruments=instruments,
+    start_date=start_date,
+    end_date=end_date,
+    initialize=initialize,
+    handle_data=handle_data,
+    order_price_field_buy='open', # 以开盘价买入
+    order_price_field_sell='open', # 以开盘价卖出
+    capital_base=100000, # 本金
+    m_deps=np.random.randn()
+    )
\ No newline at end of file