复制链接

克隆策略

In [3]:

# 本代码由可视化策略环境自动生成 2022年3月24日 11:23
# 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。


# 回测引擎：初始化函数，只执行一次
def m15_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 = 1
    # 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 = [1]
    # 设置每只股票占用的最大资金比例
    context.max_cash_per_instrument = 1
    context.options['hold_days'] = 1

# 回测引擎：每日数据处理函数，每天执行一次
def m15_handle_data_bigquant_run(context, data):
    # 获取当前持仓
    positions = {e.symbol: p.amount * p.last_sale_price
                 for e, p in context.portfolio.positions.items()}
    
    today = data.current_dt.strftime('%Y-%m-%d')
    # 按日期过滤得到今日的预测数据
    ranker_prediction = context.ranker_prediction[
        context.ranker_prediction.date == today]
#     try:
#     #大盘风控模块，读取风控数据    
#         benckmark_risk=ranker_prediction['bm_0'].values[0]
#         if benckmark_risk > 0:
#             for instrument in positions.keys():
#                 context.order_target(context.symbol(instrument), 0)
#                 print(today,'大盘风控止损触发,全仓卖出')
#                 return
#     except:
#         print('--！')
        
    #当risk为1时，市场有风险，全部平仓，不再执行其它操作    
    # 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')]
    #cash_for_buy = min(context.portfolio.portfolio_value/2,context.portfolio.cash)
    #cash_for_buy = context.portfolio.portfolio_value
    #print(ranker_prediction)
    #cash_for_buy = context.portfolio.portfolio_value
    # 手上有的资金全拿来买股票
    cash_for_buy = context.portfolio.cash
    # stockranker给出的股票排序
    buy_instruments = list(ranker_prediction.instrument)
    # 需要卖出的股票（持仓的股票）
    sell_instruments = [instrument.symbol for instrument in context.portfolio.positions.keys()]
    # 买入的股票不能是需要卖出的股票
    to_buy = set(buy_instruments[:1]) - set(sell_instruments) 
    # 需要卖出的股票不包含需要买入的股票
    to_sell = set(sell_instruments) -  set(buy_instruments[:1])
   
    
    for instrument in to_sell:
        context.order_target(context.symbol(instrument), 0)
    for instrument in to_buy:
        context.order_value(context.symbol(instrument), cash_for_buy)

def m15_prepare_bigquant_run(context):


     # 获取st状态和涨跌停状态
    
    context.status_df = D.features(instruments =context.instruments,start_date = context.start_date, end_date = context.end_date, 
                           fields=['st_status_0','price_limit_status_0','price_limit_status_1'])

def m15_before_trading_start_bigquant_run(context, data):
    pass     
#     # 获取涨跌停状态数据
#     df_price_limit_status=context.status_df.set_index('date')
#     today=data.current_dt.strftime('%Y-%m-%d')
#     # 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)
#             try:
#                 #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_0.loc[today]>2 and _order.amount<0:
#                     cancel_order(_order)
#                     print(today,'尾盘涨停取消卖单',ins) 
#             except:
#                 continue
  
    
    

m1 = M.input_features.v1(
    features="""
# #号开始的表示注释，注释需单独一行
# 多个特征，每行一个，可以包含基础特征和衍生特征，特征须为本平台特征
ts_min(amount_0,20)/mean(amount_0,20)
#20日内最低成交额/20日平均成交额
rank_swing_volatility_5_0
#5日的振幅波动率排名
rank(mean(mf_net_amount_xl_0,5))/rank(mean(mf_net_amount_xl_0,20))
#5日的平均超大单资金流入的排名/（20日的平均超大单资金流入的排名）
rank(sum(high_0/close_0,20))/rank(sum(close_0/low_0,10))
#最高价和最低价的关系
mean(mf_net_amount_m_0,10)/mean(mf_net_amount_m_0,20)
#10日内的资金流中单净值/20日内中单净值 排序
rank(mean(amount_0/deal_number_0,5))/rank(mean(amount_0/deal_number_0,20))
#成交额和成交笔数的关系
rank(mean(mf_net_amount_s_0,5))/rank(mean(mf_net_amount_s_0,20))
#5日内的资金流小单净值/20日内小单净值 排序
rank(mean(mf_net_amount_m_0,5))/rank(mean(mf_net_amount_m_0,10))
#5日内的资金流中单净值/10日内小单净值 排序
rank(mean(mf_net_amount_l_0,5))/rank(mean(mf_net_amount_l_0,10))
#5日内的资金流大单净值/10日内大单净值 排序
correlation(sqrt(volume_0),return_0,5)
#5日内收益率和成交量的平方的相关系数
correlation(log(volume_0),abs(return_0-1),5)
#5日内成交量的对数 和收益率的绝对值 的相关系数

"""
)

m2 = M.input_features.v1(
    features_ds=m1.data,
    features="""
# #号开始的表示注释，注释需单独一行
# 多个特征，每行一个，可以包含基础特征和衍生特征，特征须为本平台特征
cond1=rank(((close_0-open_0)/open_0)/((close_0-open_4)/open_4))

cond3=rank(((high_0-low_0)/close_1)/ts_max(((high_0-low_0)/close_1), 20))
"""
)

m4 = M.instruments.v2(
    start_date='2011-10-20',
    end_date='2021-03-15',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m3 = M.advanced_auto_labeler.v2(
    instruments=m4.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)-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30个分类

all_wbins(label, 30)

""",
    start_date='',
    end_date='',
    benchmark='000300.SHA',
    drop_na_label=True,
    cast_label_int=True,
    user_functions={}
)

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

m6 = 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={}
)

m8 = M.join.v3(
    data1=m3.data,
    data2=m6.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m7 = M.chinaa_stock_filter.v1(
    input_data=m8.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=['退市'],
    output_left_data=False
)

m9 = M.filter.v3(
    input_data=m7.data,
    expr='cond1<0.01 & cond3>0.85',
    output_left_data=False
)

m16 = M.dropnan.v1(
    input_data=m9.data
)

m10 = M.stock_ranker_train.v6(
    training_ds=m16.data,
    features=m1.data,
    test_ds=m16.data,
    learning_algorithm='排序',
    number_of_leaves=30,
    minimum_docs_per_leaf=280,
    number_of_trees=21,
    learning_rate=0.1,
    max_bins=1023,
    feature_fraction=1,
    data_row_fraction=1,
    plot_charts=True,
    ndcg_discount_base=1,
    m_lazy_run=False
)

m12 = M.instruments.v2(
    start_date='2021-03-16',
    end_date='2022-03-23',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m13 = M.general_feature_extractor.v7(
    instruments=m12.data,
    features=m2.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m14 = M.derived_feature_extractor.v3(
    input_data=m13.data,
    features=m2.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False,
    user_functions={}
)

m17 = M.filter.v3(
    input_data=m14.data,
    expr='cond1<0.01 & cond3>0.85',
    output_left_data=False
)

m18 = M.dropnan.v1(
    input_data=m17.data
)

m11 = M.stock_ranker_predict.v5(
    model=m10.model,
    data=m18.data,
    m_lazy_run=False
)

m15 = M.trade.v4(
    instruments=m12.data,
    options_data=m11.predictions,
    start_date='',
    end_date='',
    initialize=m15_initialize_bigquant_run,
    handle_data=m15_handle_data_bigquant_run,
    prepare=m15_prepare_bigquant_run,
    before_trading_start=m15_before_trading_start_bigquant_run,
    volume_limit=0,
    order_price_field_buy='open',
    order_price_field_sell='close',
    capital_base=100000,
    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='000300.SHA'
)

[2022-03-24 11:13:36.006021] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-03-24 11:13:36.021322] INFO: moduleinvoker: 命中缓存

[2022-03-24 11:13:36.023195] INFO: moduleinvoker: input_features.v1 运行完成[0.017186s].

[2022-03-24 11:13:36.029247] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-03-24 11:13:36.037518] INFO: moduleinvoker: 命中缓存

[2022-03-24 11:13:36.039802] INFO: moduleinvoker: input_features.v1 运行完成[0.010556s].

[2022-03-24 11:13:36.046165] INFO: moduleinvoker: instruments.v2 开始运行..

[2022-03-24 11:13:36.137261] INFO: moduleinvoker: instruments.v2 运行完成[0.091081s].

[2022-03-24 11:13:36.146808] INFO: moduleinvoker: advanced_auto_labeler.v2 开始运行..

[2022-03-24 11:13:41.608925] INFO: 自动标注(股票): 加载历史数据: 6607598 行

[2022-03-24 11:13:41.610576] INFO: 自动标注(股票): 开始标注 ..

[2022-03-24 11:13:46.642964] INFO: moduleinvoker: advanced_auto_labeler.v2 运行完成[10.496142s].

[2022-03-24 11:13:46.658550] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2022-03-24 11:13:48.596931] INFO: 基础特征抽取: 年份 2011, 特征行数=239358

[2022-03-24 11:13:50.874280] INFO: 基础特征抽取: 年份 2012, 特征行数=565675

[2022-03-24 11:13:53.217457] INFO: 基础特征抽取: 年份 2013, 特征行数=564168

[2022-03-24 11:13:55.805752] INFO: 基础特征抽取: 年份 2014, 特征行数=569948

[2022-03-24 11:13:58.633897] INFO: 基础特征抽取: 年份 2015, 特征行数=569698

[2022-03-24 11:14:01.600814] INFO: 基础特征抽取: 年份 2016, 特征行数=641546

[2022-03-24 11:14:04.881847] INFO: 基础特征抽取: 年份 2017, 特征行数=743233

[2022-03-24 11:14:08.636743] INFO: 基础特征抽取: 年份 2018, 特征行数=816987

[2022-03-24 11:14:12.963119] INFO: 基础特征抽取: 年份 2019, 特征行数=884867

[2022-03-24 11:14:17.258675] INFO: 基础特征抽取: 年份 2020, 特征行数=945961

[2022-03-24 11:14:20.523228] INFO: 基础特征抽取: 年份 2021, 特征行数=191103

[2022-03-24 11:14:20.645090] INFO: 基础特征抽取: 总行数: 6732544

[2022-03-24 11:14:20.651480] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[33.992973s].

[2022-03-24 11:14:20.662566] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2022-03-24 11:14:39.595238] INFO: derived_feature_extractor: 提取完成 cond1=rank(((close_0-open_0)/open_0)/((close_0-open_4)/open_4)), 4.802s

[2022-03-24 11:14:49.813371] INFO: derived_feature_extractor: 提取完成 cond3=rank(((high_0-low_0)/close_1)/ts_max(((high_0-low_0)/close_1), 20)), 10.217s

[2022-03-24 11:15:01.204690] INFO: derived_feature_extractor: 提取完成 ts_min(amount_0,20)/mean(amount_0,20), 11.390s

[2022-03-24 11:15:21.674178] INFO: derived_feature_extractor: 提取完成 rank(mean(mf_net_amount_xl_0,5))/rank(mean(mf_net_amount_xl_0,20)), 20.468s

[2022-03-24 11:15:42.581212] INFO: derived_feature_extractor: 提取完成 rank(sum(high_0/close_0,20))/rank(sum(close_0/low_0,10)), 20.905s

[2022-03-24 11:15:53.905023] INFO: derived_feature_extractor: 提取完成 mean(mf_net_amount_m_0,10)/mean(mf_net_amount_m_0,20), 11.322s

[2022-03-24 11:16:13.063165] INFO: derived_feature_extractor: 提取完成 rank(mean(amount_0/deal_number_0,5))/rank(mean(amount_0/deal_number_0,20)), 19.157s

[2022-03-24 11:16:33.751154] INFO: derived_feature_extractor: 提取完成 rank(mean(mf_net_amount_s_0,5))/rank(mean(mf_net_amount_s_0,20)), 20.686s

[2022-03-24 11:16:53.251609] INFO: derived_feature_extractor: 提取完成 rank(mean(mf_net_amount_m_0,5))/rank(mean(mf_net_amount_m_0,10)), 19.499s

[2022-03-24 11:17:14.175676] INFO: derived_feature_extractor: 提取完成 rank(mean(mf_net_amount_l_0,5))/rank(mean(mf_net_amount_l_0,10)), 20.923s

[2022-03-24 11:17:49.620293] INFO: derived_feature_extractor: 提取完成 correlation(sqrt(volume_0),return_0,5), 35.443s

[2022-03-24 11:18:23.486160] INFO: derived_feature_extractor: 提取完成 correlation(log(volume_0),abs(return_0-1),5), 33.864s

[2022-03-24 11:18:25.461935] INFO: derived_feature_extractor: /y_2011, 239358

[2022-03-24 11:18:26.727039] INFO: derived_feature_extractor: /y_2012, 565675

[2022-03-24 11:18:28.443307] INFO: derived_feature_extractor: /y_2013, 564168

[2022-03-24 11:18:30.134084] INFO: derived_feature_extractor: /y_2014, 569948

[2022-03-24 11:18:31.884824] INFO: derived_feature_extractor: /y_2015, 569698

[2022-03-24 11:18:33.981026] INFO: derived_feature_extractor: /y_2016, 641546

[2022-03-24 11:18:36.323685] INFO: derived_feature_extractor: /y_2017, 743233

[2022-03-24 11:18:38.912859] INFO: derived_feature_extractor: /y_2018, 816987

[2022-03-24 11:18:41.585982] INFO: derived_feature_extractor: /y_2019, 884867

[2022-03-24 11:18:44.637412] INFO: derived_feature_extractor: /y_2020, 945961

[2022-03-24 11:18:46.381302] INFO: derived_feature_extractor: /y_2021, 191103

[2022-03-24 11:18:46.818621] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[266.156034s].

[2022-03-24 11:18:46.830907] INFO: moduleinvoker: join.v3 开始运行..

[2022-03-24 11:19:00.995825] INFO: join: /y_2011, 行数=114418/239358, 耗时=3.395916s

[2022-03-24 11:19:05.690392] INFO: join: /y_2012, 行数=565669/565675, 耗时=4.691855s

[2022-03-24 11:19:10.961248] INFO: join: /y_2013, 行数=564160/564168, 耗时=5.264961s

[2022-03-24 11:19:15.755390] INFO: join: /y_2014, 行数=569942/569948, 耗时=4.787794s

[2022-03-24 11:19:22.051958] INFO: join: /y_2015, 行数=569688/569698, 耗时=6.289373s

[2022-03-24 11:19:27.616182] INFO: join: /y_2016, 行数=641542/641546, 耗时=5.558144s

[2022-03-24 11:19:33.756994] INFO: join: /y_2017, 行数=743225/743233, 耗时=6.134613s

[2022-03-24 11:19:40.679463] INFO: join: /y_2018, 行数=816971/816987, 耗时=6.912941s

[2022-03-24 11:19:47.399373] INFO: join: /y_2019, 行数=884838/884867, 耗时=6.713069s

[2022-03-24 11:19:53.793880] INFO: join: /y_2020, 行数=945905/945961, 耗时=6.386958s

[2022-03-24 11:19:57.414650] INFO: join: /y_2021, 行数=182700/191103, 耗时=3.612437s

[2022-03-24 11:19:57.535138] INFO: join: 最终行数: 6599058

[2022-03-24 11:19:57.588179] INFO: moduleinvoker: join.v3 运行完成[70.757262s].

[2022-03-24 11:19:57.606011] INFO: moduleinvoker: chinaa_stock_filter.v1 开始运行..

[2022-03-24 11:19:58.435007] INFO: A股股票过滤: 过滤 /y_2011, 642/0/114418

[2022-03-24 11:20:01.162211] INFO: A股股票过滤: 过滤 /y_2012, 2415/0/565669

[2022-03-24 11:20:03.732233] INFO: A股股票过滤: 过滤 /y_2013, 1644/0/564160

[2022-03-24 11:20:06.364377] INFO: A股股票过滤: 过滤 /y_2014, 985/0/569942

[2022-03-24 11:20:08.893176] INFO: A股股票过滤: 过滤 /y_2015, 297/0/569688

[2022-03-24 11:20:11.874464] INFO: A股股票过滤: 过滤 /y_2016, 28/0/641542

[2022-03-24 11:20:15.249724] INFO: A股股票过滤: 过滤 /y_2017, 27/0/743225

[2022-03-24 11:20:19.197680] INFO: A股股票过滤: 过滤 /y_2018, 112/0/816971

[2022-03-24 11:20:23.406638] INFO: A股股票过滤: 过滤 /y_2019, 265/0/884838

[2022-03-24 11:20:28.106975] INFO: A股股票过滤: 过滤 /y_2020, 239/0/945905

[2022-03-24 11:20:29.148457] INFO: A股股票过滤: 过滤 /y_2021, 66/0/182700

[2022-03-24 11:20:29.154736] INFO: A股股票过滤: 过滤完成, 6720 + 0

[2022-03-24 11:20:29.183416] INFO: moduleinvoker: chinaa_stock_filter.v1 运行完成[31.577372s].

[2022-03-24 11:20:29.194458] INFO: moduleinvoker: filter.v3 开始运行..

[2022-03-24 11:20:29.211669] INFO: filter: 使用表达式 cond1<0.01 & cond3>0.85 过滤

[2022-03-24 11:20:29.408518] INFO: filter: 过滤 /y_2011, 2/0/642

[2022-03-24 11:20:29.504818] INFO: filter: 过滤 /y_2012, 10/0/2415

[2022-03-24 11:20:29.585106] INFO: filter: 过滤 /y_2013, 1/0/1644

[2022-03-24 11:20:29.672030] INFO: filter: 过滤 /y_2014, 2/0/985

[2022-03-24 11:20:29.706472] INFO: filter: 过滤 /y_2015, 0/0/297

[2022-03-24 11:20:29.742174] INFO: filter: 过滤 /y_2016, 0/0/28

[2022-03-24 11:20:29.793387] INFO: filter: 过滤 /y_2017, 0/0/27

[2022-03-24 11:20:29.877055] INFO: filter: 过滤 /y_2018, 1/0/112

[2022-03-24 11:20:29.944836] INFO: filter: 过滤 /y_2019, 1/0/265

[2022-03-24 11:20:29.984929] INFO: filter: 过滤 /y_2020, 0/0/239

[2022-03-24 11:20:30.068204] INFO: filter: 过滤 /y_2021, 1/0/66

[2022-03-24 11:20:30.101522] INFO: moduleinvoker: filter.v3 运行完成[0.907074s].

[2022-03-24 11:20:30.114407] INFO: moduleinvoker: dropnan.v1 开始运行..

[2022-03-24 11:20:30.224137] INFO: dropnan: /y_2011, 2/2

[2022-03-24 11:20:30.281316] INFO: dropnan: /y_2012, 8/10

[2022-03-24 11:20:30.333006] INFO: dropnan: /y_2013, 1/1

[2022-03-24 11:20:30.387629] INFO: dropnan: /y_2014, 2/2

[2022-03-24 11:20:30.471441] INFO: dropnan: /y_2018, 1/1

[2022-03-24 11:20:30.522627] INFO: dropnan: /y_2019, 1/1

[2022-03-24 11:20:30.551763] INFO: dropnan: /y_2021, 0/1

[2022-03-24 11:20:30.597840] INFO: dropnan: 行数: 15/18

[2022-03-24 11:20:30.603063] INFO: moduleinvoker: dropnan.v1 运行完成[0.488638s].

[2022-03-24 11:20:30.611642] INFO: moduleinvoker: stock_ranker_train.v6 开始运行..

[2022-03-24 11:20:31.176220] INFO: StockRanker: 特征预处理 ..

[2022-03-24 11:20:31.234579] INFO: StockRanker: prepare data: training ..

[2022-03-24 11:20:31.275056] INFO: StockRanker: sort ..

[2022-03-24 11:20:31.327388] INFO: StockRanker: prepare data: test ..

[2022-03-24 11:20:31.365880] INFO: StockRanker: sort ..

[2022-03-24 11:20:31.419992] INFO: StockRanker训练: 5bef2462 准备训练: 15 行数, test: 15 rows

[2022-03-24 11:20:31.421470] INFO: StockRanker训练: AI模型训练，将在15*11=0.02万数据上对模型训练进行21轮迭代训练。预计将需要1～2分钟。请耐心等待。

[2022-03-24 11:20:31.615060] INFO: StockRanker训练: 正在训练 ..

[2022-03-24 11:20:31.665524] INFO: StockRanker训练: 任务状态: Pending

[2022-03-24 11:20:41.706884] INFO: StockRanker训练: 任务状态: Running

[2022-03-24 11:21:41.990623] INFO: StockRanker训练: 任务状态: Succeeded

[2022-03-24 11:21:42.002741] ERROR: moduleinvoker: module name: stock_ranker_train, module version: v6, trackeback: Exception: 模型训练失败：可能导致错误的原因是训练数据问题，请检查训练数据, err_code=1 (5bef2462ab2111ec80d3c6e6238ccd28)

---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
<ipython-input-3-9abed9487e05> in <module>
    218 )
    219 
--> 220 m10 = M.stock_ranker_train.v6(
    221     training_ds=m16.data,
    222     features=m1.data,

Exception: 模型训练失败：可能导致错误的原因是训练数据问题，请检查训练数据, err_code=1 (5bef2462ab2111ec80d3c6e6238ccd28)