复制链接

克隆策略

In [8]:

# 本代码由可视化策略环境自动生成 2022年6月2日 15:51
# 本代码单元只能在可视化模式下编辑。您也可以拷贝代码，粘贴到新建的代码单元或者策略，然后修改。


def m17_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    data = pd.read_pickle('/home/bigquant/work/userlib/model_rdr.csv')
    model_ds = DataSource.write_pickle(data.iloc[0].values[0])
    print(model_ds)
    return Outputs(data_1=model_ds, data_2=None, data_3=None, data_4=None)
# 后处理函数，可选。输入是主函数的输出，可以在这里对数据做处理，或者返回更友好的outputs数据格式。此函数输出不会被缓存。
def m17_post_run_bigquant_run(outputs):
    return outputs

# 回测引擎：初始化函数，只执行一次
def m13_initialize_bigquant_run(context):
    # 加载预测数据
    context.ranker_prediction = context.options['data'].read_df()

    # 系统已经设置了默认的交易手续费和滑点，要修改手续费可使用如下函数
    context.set_commission(PerOrder(buy_cost=0.0003, sell_cost=0.0013, min_cost=5))
    # 预测数据，通过options传入进来，使用 read_df 函数，加载到内存 (DataFrame)
    # 设置买入的股票数量，这里买入预测股票列表排名靠前的5只
    stock_count = 3
    # 每只的股票的权重，如下的权重分配会使得靠前的股票分配多一点的资金，[0.339160, 0.213986, 0.169580, ..]
    context.stock_weights = T.norm([1 / math.log(i + 2) for i in range(0, stock_count)])
    # 设置每只股票占用的最大资金比例
    context.max_cash_per_instrument = 0.2
    context.hold_days = 1
    
    from zipline.finance.slippage import SlippageModel
    class FixedPriceSlippage(SlippageModel):
        def process_order(self, data, order, bar_volume=0, trigger_check_price=0):
            if order.limit is None:
                price_field = self._price_field_buy if order.amount > 0 else self._price_field_sell
                price = data.current(order.asset, price_field)
            else:
                price = data.current(order.asset, self._price_field_buy)
            # 返回希望成交的价格和数量
            return (price, order.amount)
    # 设置price_field,默认是开盘买入，收盘卖出
    context.fix_slippage = FixedPriceSlippage(price_field_buy='open', price_field_sell='close')
    context.set_slippage(us_equities=context.fix_slippage)
    
    #大盘数据获取
    bm_df = DataSource('bar1d_index_CN_STOCK_A').read(instruments=['000001.HIX'])
    bm_df["bm_ret"] = bm_df["close"]/bm_df["close"].shift(10)-1
    bm_df["bm_ret"] = bm_df["bm_ret"].shift(1) #取昨日的收益情况
    context.bm_df = bm_df[['date','bm_ret']]
    
    #个股风控计算
    import pandas as pd
    start_date = context.ranker_prediction.date.iloc[0]
    start_date = pd.to_datetime(start_date)-pd.Timedelta(days=30)
    end_date = context.ranker_prediction.date.iloc[-1]
    stocks = context.ranker_prediction.instrument.to_list()
    data = DataSource("bar1d_CN_STOCK_A").read(instruments=stocks,start_date=start_date.strftime("%Y-%m-%d"),end_date=end_date)
    #计算个股风控，小于20日均线
    def cal_risk(df):
        df = df.sort_values("date")
        df["ma"] = df.close.rolling(20).mean()
        df["risk"] = np.where(df.close.shift(1)<df.ma.shift(1),1,0)
        return df
    context.stock_risk_data = data.groupby("instrument").apply(cal_risk).reset_index(drop=True)


# 回测引擎：每日数据处理函数，每天执行一次
def m13_handle_data_bigquant_run(context, data):
    # 按日期过滤得到今日的预测数据
    ranker_prediction = context.ranker_prediction[
        context.ranker_prediction.date == data.current_dt.strftime('%Y-%m-%d')]
    today = data.current_dt.strftime('%Y-%m-%d')
    
    # 1. 资金分配
    # 平均持仓时间是hold_days，每日都将买入股票，每日预期使用 1/hold_days 的资金
    # 实际操作中，会存在一定的买入误差，所以在前hold_days天，等量使用资金；之后，尽量使用剩余资金（这里设置最多用等量的1.5倍）
    is_staging = context.trading_day_index < context.hold_days # 是否在建仓期间（前 hold_days 天）
    cash_avg = context.portfolio.portfolio_value / context.hold_days
    cash_for_buy = min(context.portfolio.cash, (1 if is_staging else 1.5) * cash_avg)
    cash_for_sell = cash_avg - (context.portfolio.cash - cash_for_buy)
    positions = {e.symbol: p.amount * p.last_sale_price
                 for e, p in context.perf_tracker.position_tracker.positions.items()}
    equities = {e.symbol: e for e, p in context.portfolio.positions.items()}
    '''
    # 2. 生成卖出订单：hold_days天之后才开始卖出；对持仓的股票，按StockRanker预测的排序末位淘汰
    if not is_staging and cash_for_sell > 0:
        equities = {e.symbol: e for e, p in context.perf_tracker.position_tracker.positions.items()}
        instruments = list(reversed(list(ranker_prediction.instrument[ranker_prediction.instrument.apply(
                lambda x: x in equities and not context.has_unfinished_sell_order(equities[x]))])))
        # print('rank order for sell %s' % instruments)
        for instrument in instruments:
            context.order_target(context.symbol(instrument), 0)
            cash_for_sell -= positions[instrument]
            if cash_for_sell <= 0:
                break

    # 3. 生成买入订单：按StockRanker预测的排序，买入前面的stock_count只股票
    buy_cash_weights = context.stock_weights
    buy_instruments = list(ranker_prediction.instrument[:len(buy_cash_weights)])
    max_cash_per_instrument = context.portfolio.portfolio_value * context.max_cash_per_instrument
    for i, instrument in enumerate(buy_instruments):
        cash = cash_for_buy * buy_cash_weights[i]
        if cash > max_cash_per_instrument - positions.get(instrument, 0):
            # 确保股票持仓量不会超过每次股票最大的占用资金量
            cash = max_cash_per_instrument - positions.get(instrument, 0)
        if cash > 0:
            context.order_value(context.symbol(instrument), cash)
    '''
    
    # 记录持仓中st的股票
    st_stock_list = []
    name_df = context.name_df
    name_today = name_df[name_df.date==today]
    #print(name_today)
    for instrument in equities:
        name_instrument = name_today[name_today.instrument==instrument]['name'].values[0]
        # 如果股票状态变为了st 则卖出
        if 'ST' in name_instrument or '退' in name_instrument:
            # 指定一个limit_price,此时会以开盘价成交，这是由于初始化函数中改写了下单价格
            context.order_target(context.symbol(instrument), 0, limit_price=1.0)
            st_stock_list.append(instrument)
            cash_for_sell -= positions[instrument]
    if st_stock_list!=[]:
        print(today,'持仓出现st股/退市股',st_stock_list,'进行卖出处理')
 
    # 2. 生成卖出订单：hold_days天之后才开始卖出；对持仓的股票，按机器学习算法预测的排序末位淘汰
    if not is_staging and cash_for_sell > 0:
        instruments = list(reversed(list(ranker_prediction.instrument[ranker_prediction.instrument.apply(
            lambda x: x in equities)])))
        price_limit_status = context.price_limit_status
        status_today = price_limit_status[price_limit_status.date==today]
        for instrument in instruments:
            # 如果是st股票已经卖过了，就跳过
            if instrument in st_stock_list:
                continue
            # 如果涨停就跳过股票
            #status_instrument = status_today[status_today.instrument==instrument]['price_limit_status'].values[0]
            #if status_instrument>2:
            #    continue
            context.order_target(context.symbol(instrument),0)
            cash_for_sell -= positions[instrument]
            if cash_for_sell <= 0:
                break

    # 3. 生成买入订单：按机器学习算法预测的排序，买入前面的stock_count只股票
    buy_cash_weights = context.stock_weights
    buy_instruments = list(ranker_prediction.instrument[:len(buy_cash_weights)])
    pred_label = list(ranker_prediction.pred_label[:len(buy_cash_weights)])
    print(today,'今日选中标的为：')   
    print(pred_label)
    print(buy_instruments)
    print(pred_label)
    max_cash_per_instrument = context.portfolio.portfolio_value * context.max_cash_per_instrument
    for i, instrument in enumerate(buy_instruments):
        cash = cash_for_buy * buy_cash_weights[i]
        if cash > max_cash_per_instrument - positions.get(instrument, 0):
            # 确保股票持仓量不会超过每次股票最大的占用资金量
            cash = max_cash_per_instrument - positions.get(instrument, 0)
        if cash > 0:
            context.order_value(context.symbol(instrument), cash)



# 回测引擎：准备数据，只执行一次
def m13_prepare_bigquant_run(context):
    # 获取股票名称 用于过滤st和退市股
    context.name_df = DataSource('instruments_CN_STOCK_A').read()
    # 获取涨跌停状态
    context.price_limit_status = DataSource('stock_status_CN_STOCK_A').read(fields=['price_limit_status'])

from zipline.finance.order import Order

#插入定单
def insert_order(context,date,instr,amount):
    order = Order(
        dt = pd.to_datetime(date+" 09:30:00"),
        asset=context.symbol(instr),
        amount=-amount,
        stop=None,
        limit=None,
        price_field='open')

    try:
        context.blotter.open_orders[order.asset].append(order)
    except Exception:
        context.blotter.open_orders[order.asset] = [order]

    context.blotter.orders[order.id] = order
    context.blotter.new_orders.append(order) 

#个股风控判断
def stock_risk(context, data):
    today=data.current_dt.strftime('%Y-%m-%d')
    #====卖出股票
    stock_hold_now = {e.symbol:p.amount for e, p in context.perf_tracker.position_tracker.positions.items()}
    stocks = stock_hold_now.keys()

    for instr,amount in stock_hold_now.items():
        nowdata = context.stock_risk_data[(context.stock_risk_data.instrument==instr)&(context.stock_risk_data.date==today)]
        #触发个股风控，早盘卖出
        if nowdata.risk.iloc[0] == 1 and amount>0:
            print(today,'个股风控卖出：',instr) 
            insert_order(context,today,instr,amount)


def m13_before_trading_start_bigquant_run(context, data):
    df_price_limit_status=context.price_limit_status.set_index('date')
    today=data.current_dt.strftime('%Y-%m-%d')
    
    now_bm = context.bm_df[context.bm_df.date==today]
    #个股风控
    #stock_risk(context,data)
    context.bm_risk = 0
    #大盘风控判断
    '''
    if(now_bm.bm_ret.iloc[0]<-0.01):
        context.bm_risk = 1
        # 得到当前未完成订单
        for orders in get_open_orders().values():
            # 循环，撤销订单
            for _order in orders:
                ins=str(_order.sid.symbol)
                if data.can_trade(_order.sid) and _order.amount>0:
                    #大盘风控取消买单
                    cancel_order(_order)
                    print(today,'大盘风控取消买单',ins) 
                if data.can_trade(_order.sid) and _order.amount<0:#卖单由后续统一处理，先取消
                    #大盘风控取消卖单
                    cancel_order(_order)
                    print(today,'大盘风控取消卖单',ins) 
                    
        #====卖出股票
        stock_hold_now = {e.symbol:p.amount for e, p in context.perf_tracker.position_tracker.positions.items()}
        print(today,"=======风控卖出所有的股票:",stock_hold_now)
        for instr,amount in stock_hold_now.items():
            #插入定单
            insert_order(context,today,instr,amount)
    '''
    
    
    # 得到当前未完成订单
    for orders in get_open_orders().values():
        # 循环，撤销订单
        for _order in orders:
            ins=str(_order.sid.symbol)
            if data.can_trade(_order.sid):
                #判断一下如果当日涨停，则取消卖单
                if  df_price_limit_status[df_price_limit_status.instrument==ins].price_limit_status.loc[today]>2 and _order.amount<0:
                    cancel_order(_order)
                    print(today,'取消昨日尾盘涨停的卖单',ins)    
                    
     

m1 = M.instruments.v2(
    start_date='2020-01-01',
    end_date='2020-12-31',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m8 = M.advanced_auto_labeler.v2(
    instruments=m1.data,
    label_expr="""# #号开始的表示注释
# 0. 每行一个，顺序执行，从第二个开始，可以使用label字段
# 1. 可用数据字段见 https://bigquant.com/docs/develop/datasource/deprecated/history_data.html
#   添加benchmark_前缀，可使用对应的benchmark数据
# 2. 可用操作符和函数见 `表达式引擎 <https://bigquant.com/docs/develop/bigexpr/usage.html>`_

# 计算收益：2日收盘价(作为卖出价格)除以明日开盘价(作为买入价格)
shift(close, -2) / shift(open, -1)


# 极值处理：用1%和99%分位的值做clip
clip(label, all_quantile(label, 0.01), all_quantile(label, 0.99))

# 将分数映射到分类，这里使用20个分类
all_wbins(label, 2)
# 过滤掉一字涨停的情况 (设置label为NaN，在后续处理和训练中会忽略NaN的label)
#where(shift(high, -1) == shift(low, -1), NaN, label)

shift(price_limit_status,-1)>2""",
    start_date='',
    end_date='',
    benchmark='000300.SHA',
    drop_na_label=True,
    cast_label_int=False,
    user_functions={},
    m_cached=False
)

m2 = M.input_features.v1(
    features="""# #号开始的表示注释
# 多个特征，每行一个，可以包含基础特征和衍生特征
return_5
return_10
return_20
avg_amount_0/avg_amount_5
avg_amount_5/avg_amount_20
rank_avg_amount_0/rank_avg_amount_5
rank_avg_amount_5/rank_avg_amount_10
rank_return_0
rank_return_5
rank_return_10
rank_return_0/rank_return_5
rank_return_5/rank_return_10
pe_ttm_0"""
)

m3 = M.general_feature_extractor.v7(
    instruments=m1.data,
    features=m2.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m6 = M.derived_feature_extractor.v3(
    input_data=m3.data,
    features=m2.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False,
    user_functions={}
)

m9 = M.join.v3(
    data1=m8.data,
    data2=m6.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m15 = M.chinaa_stock_filter.v1(
    input_data=m9.data,
    index_constituent_cond=['全部'],
    board_cond=['上证主板', '深证主板'],
    industry_cond=['全部'],
    st_cond=['正常'],
    delist_cond=['非退市'],
    output_left_data=False
)

m11 = M.dropnan.v1(
    input_data=m15.data
)

m4 = M.instruments.v2(
    start_date='2022-01-01',
    end_date='2022-02-01',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m5 = M.general_feature_extractor.v7(
    instruments=m4.data,
    features=m2.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m7 = M.derived_feature_extractor.v3(
    input_data=m5.data,
    features=m2.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False,
    user_functions={}
)

m16 = M.chinaa_stock_filter.v1(
    input_data=m7.data,
    index_constituent_cond=['全部'],
    board_cond=['上证主板', '深证主板'],
    industry_cond=['全部'],
    st_cond=['正常'],
    delist_cond=['非退市'],
    output_left_data=False
)

m10 = M.dropnan.v1(
    input_data=m16.data
)

m17 = M.cached.v3(
    run=m17_run_bigquant_run,
    post_run=m17_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports=''
)

m12 = M.random_forest_regressor.v1(
    training_ds=m17.data_1,
    predict_ds=m10.data,
    iterations=10,
    feature_fraction=1,
    max_depth=30,
    min_samples_per_leaf=200,
    key_cols='date,instrument',
    workers=1,
    random_state=0,
    other_train_parameters={"random_state":0}
)

m14 = M.sort.v4(
    input_ds=m12.predictions,
    sort_by='pred_label',
    group_by='date',
    keep_columns='--',
    ascending=False
)

m13 = M.trade.v4(
    instruments=m4.data,
    options_data=m14.sorted_data,
    start_date='',
    end_date='',
    initialize=m13_initialize_bigquant_run,
    handle_data=m13_handle_data_bigquant_run,
    prepare=m13_prepare_bigquant_run,
    before_trading_start=m13_before_trading_start_bigquant_run,
    volume_limit=0.025,
    order_price_field_buy='open',
    order_price_field_sell='close',
    capital_base=1000000,
    auto_cancel_non_tradable_orders=True,
    data_frequency='daily',
    price_type='真实价格',
    product_type='股票',
    plot_charts=True,
    backtest_only=False,
    benchmark=''
)

[2022-06-02 15:31:53.189905] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-06-02 15:31:53.293987] INFO: moduleinvoker: 命中缓存

[2022-06-02 15:31:53.318070] INFO: moduleinvoker: input_features.v1 运行完成[0.120591s].

[2022-06-02 15:31:53.366678] INFO: moduleinvoker: instruments.v2 开始运行..

[2022-06-02 15:31:53.384074] INFO: moduleinvoker: 命中缓存

[2022-06-02 15:31:53.386487] INFO: moduleinvoker: instruments.v2 运行完成[0.019838s].

[2022-06-02 15:31:53.465842] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2022-06-02 15:31:53.479729] INFO: moduleinvoker: 命中缓存

[2022-06-02 15:31:53.487440] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[0.021614s].

[2022-06-02 15:31:53.499178] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2022-06-02 15:31:53.517477] INFO: moduleinvoker: 命中缓存

[2022-06-02 15:31:53.520012] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[0.020835s].

[2022-06-02 15:31:53.535264] INFO: moduleinvoker: chinaa_stock_filter.v1 开始运行..

[2022-06-02 15:31:53.550787] INFO: moduleinvoker: 命中缓存

[2022-06-02 15:31:53.553306] INFO: moduleinvoker: chinaa_stock_filter.v1 运行完成[0.018039s].

[2022-06-02 15:31:53.572220] INFO: moduleinvoker: dropnan.v1 开始运行..

[2022-06-02 15:31:53.580314] INFO: moduleinvoker: 命中缓存

[2022-06-02 15:31:53.583168] INFO: moduleinvoker: dropnan.v1 运行完成[0.010982s].

[2022-06-02 15:31:53.615995] INFO: moduleinvoker: cached.v3 开始运行..

[2022-06-02 15:31:53.739702] INFO: moduleinvoker: cached.v3 运行完成[0.123701s].

[2022-06-02 15:31:53.767977] INFO: moduleinvoker: random_forest_regressor.v1 开始运行..

[2022-06-02 15:31:53.776125] ERROR: random_forest_regressor: training_ds不为空，准备训练，请输入参数features

[2022-06-02 15:31:53.780588] ERROR: moduleinvoker: module name: random_forest_regressor, module version: v1, trackeback: Exception: 请输入参数features

DataSource(9ca1aed308fd4d0fa56037ca5c6fbf21T)

---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
<ipython-input-8-30361800f093> in <module>
     78 )
     79 
---> 80 m12 = M.random_forest_regressor.v1(
     81     training_ds=m17.data_1,
     82     predict_ds=m10.data,

Exception: 请输入参数features

In [2]:

# import pandas as pd
# print(m12)
# pd.DataFrame([m12.output_model]).to_pickle('/home/bigquant/work/userlib/model_tree.csv')
# m12.feature_gains
print(1)

In [3]:

m12.predictions.read()

Out[3]:

	pred_label	date	instrument
0	0.015236	2021-10-08	000001.SZA
1	0.023202	2021-10-11	000001.SZA
2	0.016433	2021-10-12	000001.SZA
3	0.015236	2021-10-13	000001.SZA
4	0.010896	2021-10-14	000001.SZA
...	...	...	...
359892	0.010896	2022-01-24	605599.SHA
359893	0.028886	2022-01-25	605599.SHA
359894	0.015415	2022-01-26	605599.SHA
359895	0.010896	2022-01-27	605599.SHA
359896	0.015236	2022-01-28	605599.SHA

237012 rows × 3 columns