复制链接

克隆策略

In [22]:

# #m7.data.read_df().columns
# #m17.predictions.read_df()
# #m8.predictions.read_df().head(100)
# #m8.predictions.read_df().tail(100)
# #m2.data_2.read_df()
# #m23.data_1.read_df()
# #m23.data_2.read_df()
# # df = m11.data_1.read()['data']
# # print(df[df['date']=='2021-01-26'])
# # print(df[df['date']=='2021-01-27'])

# # 计算n日涨幅
# def zhangfu(data, timeperiod=13):          
#     adj_close = data['close']- data['close'].shift(timeperiod)             
#     return adj_close

# # 计算动量
# def barAdjMon(data, timeperiod=2):          
#     adj_close = (data['close'] + data['high'] + data['low']) / 3               
#     return np.log(adj_close / adj_close.shift(timeperiod))

# def del_main(start_date_input,end_date_input):
#     instruments = ['510330.HOF','161017.ZOF','159949.ZOF']
#     df = DataSource("bar1d_CN_FUND").read(start_date=start_date_input,instruments = instruments, end_date=end_date_input)
#     groups = df.groupby(df['instrument'])
#     df_expend = pd.DataFrame()
#     for x in instruments:
#         tp = groups.get_group(x)
#         tp = tp.sort_values(by=['date'],na_position='first')
#         tp['pre_close'] = tp['close'].shift(1)
#         print(tp.head(2))
#         df_expend = df_expend.append(tp)
#     #df_159949 = groups.get_group('159949.ZOF')
#     df_expend.sort_values(['date'],na_position='first',inplace=True)
#     df_expend = df_expend.reset_index()
#     df_expend.drop('index',axis= 1,inplace = True)
    
#     #计算其他特征
#     df_expend['mom_20'] = barAdjMon(df_expend,timeperiod=20)
#     df_expend['mom_20'] = df_expend['mom_20'].fillna(value=0.0) #列向前填充
#     # 涨幅因子
#     df_expend['zhangfu'] = zhangfu(df_expend,timeperiod=13)
#     df_expend['zhangfu'] = df_expend['zhangfu'].fillna(value=0.0) #列向前填充
#     return df_expend

# # Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
# def bigquant_run(input_1, input_2, input_3,start_date_input,end_date_input,pre_start_date,pre_end_date):
#     # 示例代码如下。在这里编写您的代码
#     # 读取数据  默认会返回全部证券代码数据, 通过指定参数 instruments 可以读取到指定的证券代码数据
#     df_expend = del_main(start_date_input,end_date_input)
#     data_1 = DataSource.write_df(df_expend)
#     data_2 = DataSource.write_df(df_expend)
#     df_pre = del_main(pre_start_date,pre_end_date)
#     data_3 = DataSource.write_df(df_pre)
#     return Outputs(data_1=data_1, data_2=data_2, data_3=data_3)

In [5]:

#m18.data.read().tail(24)
# num = m18.data.read().isna().sum()
# print(type(num))
# print(num.index[0])
# for i in range(len(num)):
#     if num[i] > 10:
#         print(num.index[i])
# import empyrical
# returns = result.get('m21').read_raw_perf()['returns'].tolist()
# sharp_ratio = empyrical.sharpe_ratio(np.array(returns),0.03/252)
from itertools import combinations

# def combine(temp_list, n):
#     '''根据n获得列表中的所有可能组合（n个元素为一组）'''
#     temp_list2 = []
#     for c in combinations(temp_list, n):
#         if len(c) != 0:
#             temp_list2.append(list(c))
#     return temp_list2

# list1 = ['a', 'b', 'c', 'd', 'e', 'f']
# end_list = []
# for i in range(len(list1)):
#     end_list.extend(combine(list1, i))
# print(end_list)

[['a'], ['b'], ['c'], ['d'], ['e'], ['f'], ['a', 'b'], ['a', 'c'], ['a', 'd'], ['a', 'e'], ['a', 'f'], ['b', 'c'], ['b', 'd'], ['b', 'e'], ['b', 'f'], ['c', 'd'], ['c', 'e'], ['c', 'f'], ['d', 'e'], ['d', 'f'], ['e', 'f'], ['a', 'b', 'c'], ['a', 'b', 'd'], ['a', 'b', 'e'], ['a', 'b', 'f'], ['a', 'c', 'd'], ['a', 'c', 'e'], ['a', 'c', 'f'], ['a', 'd', 'e'], ['a', 'd', 'f'], ['a', 'e', 'f'], ['b', 'c', 'd'], ['b', 'c', 'e'], ['b', 'c', 'f'], ['b', 'd', 'e'], ['b', 'd', 'f'], ['b', 'e', 'f'], ['c', 'd', 'e'], ['c', 'd', 'f'], ['c', 'e', 'f'], ['d', 'e', 'f'], ['a', 'b', 'c', 'd'], ['a', 'b', 'c', 'e'], ['a', 'b', 'c', 'f'], ['a', 'b', 'd', 'e'], ['a', 'b', 'd', 'f'], ['a', 'b', 'e', 'f'], ['a', 'c', 'd', 'e'], ['a', 'c', 'd', 'f'], ['a', 'c', 'e', 'f'], ['a', 'd', 'e', 'f'], ['b', 'c', 'd', 'e'], ['b', 'c', 'd', 'f'], ['b', 'c', 'e', 'f'], ['b', 'd', 'e', 'f'], ['c', 'd', 'e', 'f'], ['a', 'b', 'c', 'd', 'e'], ['a', 'b', 'c', 'd', 'f'], ['a', 'b', 'c', 'e', 'f'], ['a', 'b', 'd', 'e', 'f'], ['a', 'c', 'd', 'e', 'f'], ['b', 'c', 'd', 'e', 'f']]

In [115]:

# 本代码由可视化策略环境自动生成 2022年7月25日 01:26
# 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。


from itertools import combinations

def combine(temp_list, n):
    '''根据n获得列表中的所有可能组合（n个元素为一组）'''
    temp_list2 = []
    for c in combinations(temp_list, n):
        if len(c) != 0:
            temp_list2.append(list(c))
    return temp_list2


# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m9_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    list1 = input_1.read()
    end_list = []
    for i in range(len(list1)):
        end_list.extend(combine(list1, i))
    print(end_list)
    df = end_list
    data_1 = DataSource.write_pickle(df)
    data_2 = DataSource.write_pickle(df)
    return Outputs(data_1=data_1, data_2=data_2, data_3=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 m9_post_run_bigquant_run(outputs):
    return outputs

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m14_run_bigquant_run(input_1, input_2, input_3, stock_count, hold_days):
    # 示例代码如下。在这里编写您的代码
    param = {
        "stock_count": stock_count,
        "hold_days": hold_days
        }
    data_1 = DataSource.write_pickle(param)
    return Outputs(data_1=data_1, data_2=None, data_3=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 m14_post_run_bigquant_run(outputs):
    return outputs

from scipy import nanstd
from sklearn import preprocessing
import math
import numpy as np
def print_debug(**kwargs):
    if True:
        for k,v in kwargs.items():
            print(k,":")
            print(v)
            
# 计算n日涨幅
def zhangfu(data, timeperiod=13):          
    adj_close = data['pre_close']- data['pre_close'].shift(timeperiod)             
    return adj_close
def rank_swing_volatility(data,timeperiod=5):
    return nanstd((data['pre_high']-data['pre_low'])/data['pre_close'], timeperiod)*sqrt(200)*100
# 计算动量
def barAdjMon(data, timeperiod=2):          
    adj_close = (data['pre_close'] + data['pre_high'] + data['pre_low']) / 3               
    return np.log(adj_close / adj_close.shift(timeperiod))

def del_main(start_date_input,end_date_input,instruments):
    #instruments = ['510330.HOF','161017.ZOF','159949.ZOF']
    df = DataSource("bar1d_CN_FUND").read(start_date=start_date_input,instruments = instruments, end_date=end_date_input)
    groups = df.groupby(df['instrument'])
    df_expend = pd.DataFrame()
    for x in instruments:
        #分组排序后重新设置索引
        tp = groups.get_group(x)
        tp = tp.sort_values(by=['date'],na_position='first')

        tp = tp.reset_index() 
        tp.drop('index',axis= 1,inplace = True) 
        tp = tp.reset_index() 
        print_debug(tp_index=tp.index)
        # 基础特征列表
        tp['open_0'] = tp['open']
        tp['high_0'] = tp['high']
        tp['low_0'] = tp['low']
        tp['close_0'] = tp['close']
        tp['volume_0'] = tp['volume']
        tp['pre_close'] = tp['close'].shift(1)
        tp['std_pre_close'] = preprocessing.scale(tp['pre_close'].values)
        tp['pre_high'] = tp['high'].shift(1)
        tp['std_pre_high'] = preprocessing.scale(tp['pre_high'].values)
        tp['pre_low'] = tp['low'].shift(1)
        tp['std_pre_low'] = preprocessing.scale(tp['pre_low'].values)
        tp['pre_open'] = tp['open'].shift(1)
        tp['std_pre_open'] = preprocessing.scale(tp['pre_open'].values)
        tp['pre_volume'] = tp['volume'].shift(1)
        tp['std_pre_volume'] = preprocessing.scale(tp['pre_volume'].values)
        tp['pre_amount'] = tp['amount'].shift(1)
        tp['std_pre_amount'] = preprocessing.scale(tp['pre_amount'].values)
        #计算其他特征
        #tp['alpha_001_my'] = (rank(ts_argmax(signedpower(where(((tp['close_0']/shift(tp['close_0'],1)-1) < 0), std((tp['close_0']/shift(tp['close_0'],1)-1), 20), tp['close_0']), 2), 5)) -0.5)
        #20日动量
        tp['mom_20'] = barAdjMon(tp,timeperiod=20)
        tp['std_mom_20'] = preprocessing.scale(tp['mom_20'].values)
        #13日涨幅
        tp['zhangfu'] = zhangfu(tp,timeperiod=13)
        tp['std_zhangfu'] = preprocessing.scale(tp['zhangfu'].values)
        # std ln pre open
        tp['std_ln_pre_open'] = preprocessing.scale(np.log(tp['pre_open']))
        print(tp.head(2))
        df_expend = df_expend.append(tp)
    #df_159949 = groups.get_group('159949.ZOF')
    df_expend.sort_values(['date'],na_position='first',inplace=True)
    df_expend = df_expend.reset_index()
    df_expend.drop('index',axis= 1,inplace = True)
    # 异常数据处理
    #df_expend = df_expend.fillna(value=0.0) #列向前填充
#     # 波动率因子
#     df_expend['rank_swing_volatility_5'] = rank_swing_volatility(df_expend,5)
#     df_expend['rank_swing_volatility_5'] = df_expend['rank_swing_volatility_5'].fillna(value=0.0) #列向前填充
    return df_expend

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m2_run_bigquant_run(input_1, input_2, input_3,start_date_input,end_date_input,pre_start_date,pre_end_date):
    pre_start_date = input_1.read()['start_date']
    pre_end_date = input_1.read()['end_date']
    instruments = input_1.read()['instruments']
    # 示例代码如下。在这里编写您的代码
    # 读取数据  默认会返回全部证券代码数据, 通过指定参数 instruments 可以读取到指定的证券代码数据
    df_expend = del_main(start_date_input,end_date_input,instruments)
    data_1 = DataSource.write_df(df_expend)
    data_2 = DataSource.write_df(df_expend)
    print('pre_start_date,pre_end_date',pre_start_date,pre_end_date)
    df_pre = del_main(pre_start_date,pre_end_date,instruments)
    data_3 = DataSource.write_df(df_pre)
    return Outputs(data_1=data_1, data_2=data_2, data_3=data_3)

# 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 m2_post_run_bigquant_run(outputs):
    return outputs

def del_input(input_1):
    df = input_1.read_df()
    data_1 = DataSource.write_df(df)
    return data_1
# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m23_run_bigquant_run(input_1, input_2, input_3):
    data_1 = del_input(input_1)
    data_2 = del_input(input_2)
    return Outputs(data_1=data_1, data_2=data_2, data_3=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 m23_post_run_bigquant_run(outputs):
    return outputs

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def del_dirty_data(df):
    # 异常数据处理
    num = df.isna().sum()
    for i in range(len(num)):
        if num[i] > 10:
            print('!!!!!!!!!含有大量脏数据',num.index[i],num[i])
        if num[i] > 100:
            print(num.index[i],num[i])
            df = df.drop([str(num.index[i])],axis = 1)
    # 删除值全为nan的列
    df = df.dropna(axis = 1,how = 'all')
    # 删除任意含有nan的行
    #df = df.dropna()
    df = df.fillna(value=0.0) #列向前填充  
    return df
def m1_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    df1 = input_1.read()
    df1 = del_dirty_data(df1)
    data_1 = DataSource.write_df(df1)
    
    df2 = input_2.read()
    df2 = del_dirty_data(df2)
    data_2 = DataSource.write_df(df2)    
    return Outputs(data_1=data_1, data_2=data_2, data_3=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 m1_post_run_bigquant_run(outputs):
    return outputs

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
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

    print(input_3.read())
    train_df = input_1.read()
    predict_df = input_2.read()
    features = input_3.read()
    data_1 = input_1
    data_2 = input_2
    for i in features:
        print(i)
    
    return Outputs(data_1=data_1, data_2=data_2, data_3=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

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m11_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    print('m11.input1',input_1)
    df = input_1.read()
    param = input_2.read()
    
    data = {
        "param": param,
        "data": df
    }
    data_1 = DataSource.write_pickle(data)
    return Outputs(data_1=data_1, data_2=None, data_3=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 m11_post_run_bigquant_run(outputs):
    return outputs

# 回测引擎：初始化函数，只执行一次
def m21_initialize_bigquant_run(context):
    #print('初始化函数开始：')
    #print('context.options',context.options)
    # 加载预测数据
    context.ranker_prediction = context.options.get('data').read()['data']
    context.param = context.options['data'].read()["param"]
    # 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=0.0003, 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 = context.param["stock_count"]
    # 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 = 1.01
    context.hold_days = context.param["hold_days"]
    #print('初始化函数结束')

# 交易引擎：每个单位时间开盘前调用一次。
def m21_before_trading_start_bigquant_run(context, data):
    # 盘前处理，订阅行情等

    context.subscribe(context.instruments)

    pass

# 交易引擎：tick数据处理函数，每个tick执行一次
def m21_handle_tick_bigquant_run(context, tick):
    pass

import math
# 回测引擎：每日数据处理函数，每天执行一次
def m21_handle_data_bigquant_run(context, data):
    #print('当前日期data.current_dt',data.current_dt)
    #print('data',data)
    try:
        context.ranker_prediction = context.options.get('data').read()['data']
        # 相隔几天（hold_days）进行一下换仓
        if context.trading_day_index % context.hold_days != 0:
            return 

        # 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')]
        # 目前持仓
        positions = {e: p.amount * p.last_sale_price for e, p in context.portfolio.positions.items()}
        # 权重
        buy_cash_weights = context.stock_weights
        # 今日买入股票列表
        stock_to_buy = list(ranker_prediction.instrument[:len(buy_cash_weights)])
        # 持仓上限
        max_cash_per_instrument = context.portfolio.portfolio_value * context.max_cash_per_instrument
        print("<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<")
        # 通过positions对象，使用列表生成式的方法获取目前持仓的股票列表
        stock_hold_now = [equity for equity in context.portfolio.positions ]
        # 继续持有的股票：调仓时，如果买入的股票已经存在于目前的持仓里，那么应继续持有
        no_need_to_sell = [i for i in stock_hold_now if i in stock_to_buy]
        # 需要卖出的股票
        stock_to_sell = [i for i in stock_hold_now if i not in no_need_to_sell]

        # 卖出
        for stock in stock_to_sell:
            # 如果该股票停牌，则没法成交。因此需要用can_trade方法检查下该股票的状态
            # 如果返回真值，则可以正常下单，否则会出错
            # 因为stock是字符串格式，我们用symbol方法将其转化成平台可以接受的形式：Equity格式
            if data.can_trade(context.symbol(stock)):
                # order_target_percent是平台的一个下单接口，表明下单使得该股票的权重为0，
                #   即卖出全部股票，可参考回测文档
                #print('卖出order target percent',context.symbol(stock))
                #print('卖出结果',context.order_target_percent(context.symbol(stock), 0))
                context.order_target_percent(context.symbol(stock), 0)

        # 如果当天没有买入的股票，就返回
        if len(stock_to_buy) == 0:
            return

        # 买入
        print('买入列表',stock_to_buy)
        for i, instrument in enumerate(stock_to_buy):
            cash = context.portfolio.portfolio_value * 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 > 500:
                cash = int(math.floor(cash))
                #print('买入order_value',context.symbol(instrument),' cash ',cash)
                #print(context.order_value(context.symbol(instrument), cash))
                context.order_value(context.symbol(instrument), cash)
    except Exception as e:
        print('抛出异常',e)
# 交易引擎：成交回报处理函数，每个成交发生时执行一次
def m21_handle_trade_bigquant_run(context, trade):
    pass

# 交易引擎：委托回报处理函数，每个委托变化时执行一次
def m21_handle_order_bigquant_run(context, order):
    pass

# 交易引擎：盘后处理函数，每日盘后执行一次
def m21_after_trading_bigquant_run(context, data):
    pass

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def deal_result_data(df):
    
    return df
def m12_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    print(input_1)
    print(input_1.read())
    data_1 = input_1
    data_2 = input_1
#     df1 = input_1.read_raw_perf()
#     df1 = deal_result_data(df1)
#     data_1 = DataSource.write_df(df1)
    
#     df2 = input_2.read()
#     df2 = deal_result_data(df2)
#     data_2 = DataSource.write_df(df2)    
    return Outputs(data_1=data_1, data_2=data_2, data_3=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 m12_post_run_bigquant_run(outputs):
    return outputs


g = T.Graph({

    'm3': 'M.input_features.v1',
    'm3.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

# rank_swing_volatility_5 = nanstd((high-low)/pre_close, 5)*sqrt(200)*100
# alpha_001 = (rank(ts_argmax(signedpower(where(((close_0/shift(close_0,1)-1) < 0), std((close_0/shift(close_0,1)-1), 20), close_0), 2), 5)) -0.5)
# shift(close_0,5)/close_0

# mom_20
# zhangfu
# ln_pre_high = log(pre_high)
#ln_pre_open = nanstd(log(pre_open))
# std_ln_pre_open
# delta(close_0, 3)
pre_close
pre_high
pre_volume
# pre_amount

# high
# open
# close
# amount
# volume
# std_mom_20
# std_zhangfu
# std_pre_high
# std_pre_close
# std_pre_open
# std_pre_low
# std_pre_volume
# std_pre_amount

# ln_open = log(open_0)
""",

    'm9': 'M.cached.v3',
    'm9.input_1': T.Graph.OutputPort('m3.data'),
    'm9.run': m9_run_bigquant_run,
    'm9.post_run': m9_post_run_bigquant_run,
    'm9.input_ports': '',
    'm9.params': '{}',
    'm9.output_ports': '',

    'm14': 'M.cached.v3',
    'm14.run': m14_run_bigquant_run,
    'm14.post_run': m14_post_run_bigquant_run,
    'm14.input_ports': '',
    'm14.params': """{
    "stock_count": 1,
    "hold_days": 1 
}""",
    'm14.output_ports': '',

    'm4': 'M.instruments.v2',
    'm4.start_date': '2021-06-01',
    'm4.end_date': T.live_run_param('trading_date', '2022-06-01'),
    'm4.market': 'CN_FUND',
    'm4.instrument_list': """510330.HOF
161017.ZOF
159949.ZOF""",
    'm4.max_count': 0,

    'm2': 'M.cached.v3',
    'm2.input_1': T.Graph.OutputPort('m4.data'),
    'm2.run': m2_run_bigquant_run,
    'm2.post_run': m2_post_run_bigquant_run,
    'm2.input_ports': '',
    'm2.params': """{"start_date_input":"2020-06-01",
"end_date_input":"2021-05-01",
"pre_start_date":"2021-06-01",
"pre_end_date":"2022-06-01"}""",
    'm2.output_ports': '',
    'm2.m_cached': False,

    'm10': 'M.auto_labeler_on_datasource.v1',
    'm10.input_data': T.Graph.OutputPort('m2.data_1'),
    'm10.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
# 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：5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, 20)

# 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)
""",
    'm10.drop_na_label': True,
    'm10.cast_label_int': True,
    'm10.date_col': 'date',
    'm10.instrument_col': 'instrument',
    'm10.user_functions': {},

    'm23': 'M.cached.v3',
    'm23.input_1': T.Graph.OutputPort('m2.data_2'),
    'm23.input_2': T.Graph.OutputPort('m2.data_3'),
    'm23.run': m23_run_bigquant_run,
    'm23.post_run': m23_post_run_bigquant_run,
    'm23.input_ports': '',
    'm23.params': '{}',
    'm23.output_ports': '',
    'm23.m_cached': False,

    'm16': 'M.derived_feature_extractor.v3',
    'm16.input_data': T.Graph.OutputPort('m23.data_1'),
    'm16.features': T.Graph.OutputPort('m3.data'),
    'm16.date_col': 'date',
    'm16.instrument_col': 'instrument',
    'm16.drop_na': False,
    'm16.remove_extra_columns': False,

    'm7': 'M.join.v3',
    'm7.data1': T.Graph.OutputPort('m10.data'),
    'm7.data2': T.Graph.OutputPort('m16.data'),
    'm7.on': 'date,instrument',
    'm7.how': 'inner',
    'm7.sort': False,

    'm18': 'M.derived_feature_extractor.v3',
    'm18.input_data': T.Graph.OutputPort('m23.data_2'),
    'm18.features': T.Graph.OutputPort('m3.data'),
    'm18.date_col': 'date',
    'm18.instrument_col': 'instrument',
    'm18.drop_na': False,
    'm18.remove_extra_columns': False,

    'm1': 'M.cached.v3',
    'm1.input_1': T.Graph.OutputPort('m7.data'),
    'm1.input_2': T.Graph.OutputPort('m18.data'),
    'm1.run': m1_run_bigquant_run,
    'm1.post_run': m1_post_run_bigquant_run,
    'm1.input_ports': '',
    'm1.params': '{}',
    'm1.output_ports': '',

    'm17': 'M.cached.v3',
    'm17.input_1': T.Graph.OutputPort('m1.data_1'),
    'm17.input_2': T.Graph.OutputPort('m1.data_2'),
    'm17.input_3': T.Graph.OutputPort('m3.data'),
    'm17.run': m17_run_bigquant_run,
    'm17.post_run': m17_post_run_bigquant_run,
    'm17.input_ports': '',
    'm17.params': '{\'index_my\':0}',
    'm17.output_ports': '',

    'm6': 'M.stock_ranker_train.v6',
    'm6.training_ds': T.Graph.OutputPort('m17.data_1'),
    'm6.features': T.Graph.OutputPort('m3.data'),
    'm6.learning_algorithm': '排序',
    'm6.number_of_leaves': 3,
    'm6.minimum_docs_per_leaf': 100,
    'm6.number_of_trees': 20,
    'm6.learning_rate': 0.1,
    'm6.max_bins': 1023,
    'm6.feature_fraction': 1,
    'm6.data_row_fraction': 1,
    'm6.plot_charts': True,
    'm6.ndcg_discount_base': 1,
    'm6.m_lazy_run': False,
    'm6.m_cached': False,

    'm8': 'M.stock_ranker_predict.v5',
    'm8.model': T.Graph.OutputPort('m6.model'),
    'm8.data': T.Graph.OutputPort('m17.data_2'),
    'm8.m_lazy_run': False,

    'm11': 'M.cached.v3',
    'm11.input_1': T.Graph.OutputPort('m8.predictions'),
    'm11.input_2': T.Graph.OutputPort('m14.data_1'),
    'm11.run': m11_run_bigquant_run,
    'm11.post_run': m11_post_run_bigquant_run,
    'm11.input_ports': '',
    'm11.params': '{}',
    'm11.output_ports': '',

    'm21': 'M.hftrade.v2',
    'm21.instruments': T.Graph.OutputPort('m4.data'),
    'm21.options_data': T.Graph.OutputPort('m11.data_1'),
    'm21.start_date': '',
    'm21.end_date': '',
    'm21.initialize': m21_initialize_bigquant_run,
    'm21.before_trading_start': m21_before_trading_start_bigquant_run,
    'm21.handle_tick': m21_handle_tick_bigquant_run,
    'm21.handle_data': m21_handle_data_bigquant_run,
    'm21.handle_trade': m21_handle_trade_bigquant_run,
    'm21.handle_order': m21_handle_order_bigquant_run,
    'm21.after_trading': m21_after_trading_bigquant_run,
    'm21.capital_base': 1000000,
    'm21.frequency': 'daily',
    'm21.price_type': '真实价格',
    'm21.product_type': '自动',
    'm21.before_start_days': '80',
    'm21.order_price_field_buy': 'open',
    'm21.order_price_field_sell': 'close',
    'm21.benchmark': '000300.HIX',
    'm21.plot_charts': True,
    'm21.disable_cache': False,
    'm21.replay_bdb': False,
    'm21.show_debug_info': True,
    'm21.backtest_only': False,

    'm12': 'M.cached.v3',
    'm12.input_1': T.Graph.OutputPort('m21.raw_perf'),
    'm12.run': m12_run_bigquant_run,
    'm12.post_run': m12_post_run_bigquant_run,
    'm12.input_ports': '',
    'm12.params': '{}',
    'm12.output_ports': '',
})

# g.run({})


def m5_param_grid_builder_bigquant_run():
    param_grid = {}
    # 在这里设置需要调优的参数备选
    # param_grid['m3.features'] = ['close_1/close_0', 'close_2/close_0\nclose_3/close_0']
    # param_grid['m6.number_of_trees'] = [5, 10, 20]
    param_grid["m17.params"] = [
        """{"index_my": 3}""",
        """{"index_my": 4}""",
    ]
    return param_grid

def m5_scoring_bigquant_run(result):
    score = result.get('m19').read_raw_perf()['sharpe'].tail(1)[0]

    return {'score': score}


m5 = M.hyper_parameter_search.v1(
    param_grid_builder=m5_param_grid_builder_bigquant_run,
    scoring=m5_scoring_bigquant_run,
    search_algorithm='网格搜索',
    search_iterations=10,
    workers=1,
    worker_distributed_run=True,
    worker_silent=True,
    run_now=True,
    bq_graph=g
)

[2022-07-25 01:17:41.911735] INFO: moduleinvoker: 任务id: 31078, 不能继续启动AI任务, 需排队等待, 150秒后将重试..

[2022-07-25 01:20:12.110195] INFO: moduleinvoker: 31078: AI任务启动成功..

[2022-07-25 01:20:12.131150] INFO: cached.v3.dff9e194: 任务状态: Pending

[2022-07-25 01:20:22.162881] INFO: cached.v3.dff9e194: 任务状态: Running

[2022-07-25 01:20:42.223162] INFO: cached.v3.dff9e194: 任务状态: Succeeded

[2022-07-25 01:20:42.227280] ERROR: moduleinvoker: job_id: dff9e1940b7411edacd8360c1f5bd488, outputs error: 'high'

[2022-07-25 01:20:42.418245] INFO: cached.v3.f2031bf8: 任务状态: Pending

[2022-07-25 01:20:52.450692] INFO: cached.v3.f2031bf8: 任务状态: Running

[2022-07-25 01:22:22.712847] INFO: cached.v3.f2031bf8: 任务状态: Succeeded

[2022-07-25 01:22:22.716376] ERROR: moduleinvoker: job_id: f2031bf80b7411eda466360c1f5bd488, outputs error: 'high'

[2022-07-25 01:22:22.919771] INFO: cached.v3.2de94962: 任务状态: Pending

[2022-07-25 01:22:32.956454] INFO: cached.v3.2de94962: 任务状态: Running

[2022-07-25 01:23:03.059487] INFO: cached.v3.2de94962: 任务状态: Succeeded

[2022-07-25 01:23:03.066215] ERROR: moduleinvoker: job_id: 2de949620b7511ed8979360c1f5bd488, outputs error: 'high'

[2022-07-25 01:23:03.067448] INFO: moduleinvoker: hyper_parameter_search.v1 运行完成[321.363275s].

Fitting 1 folds for each of 2 candidates, totalling 2 fits
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[CV 1/1; 1/2] START m17.params={"index_my": 3}..................................

[CV 1/1; 1/2] END ................m17.params={"index_my": 3}; total time= 3.0min
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:  3.0min remaining:    0.0s
[CV 1/1; 2/2] START m17.params={"index_my": 4}..................................

[CV 1/1; 2/2] END ................m17.params={"index_my": 4}; total time= 1.7min
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:  4.7min remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:  4.7min finished