first blood

策略分享
标签: #<Tag:0x00007fc070812cd8>

(byronfitzgerald) #1

克隆策略
In [1]:
# 基础参数配置
class conf:
    start_date = '2010-01-01'
    end_date='2017-01-01'
    # split_date 之前的数据用于训练,之后的数据用作效果评估
    split_date = '2015-01-01'
    # D.instruments: https://bigquant.com/docs/data_instruments.html
    instruments = D.instruments(start_date, split_date)

    # 机器学习目标标注函数
    # 如下标注函数等价于 min(max((持有期间的收益 * 100), -20), 20) + 20 (后面的M.fast_auto_labeler会做取整操作)
    # 说明:max/min这里将标注分数限定在区间[-20, 20],+20将分数变为非负数 (StockRanker要求标注分数非负整数)
    label_expr = ['return * 100', 'where(label > {0}, {0}, where(label < -{0}, -{0}, label)) + {0}'.format(20)]
    # 持有天数,用于计算label_expr中的return值(收益)
    hold_days = 5

    # 特征 https://bigquant.com/docs/data_features.html,你可以通过表达式构造任何特征
    features = [
        'return_5',  # 5日收益
        'rank_return_5',  # 5日收益排名
        'avg_turn_5',  # 5日平均换手率
        'avg_amount_5',  # 5日平均交易额
        '(volume_0+volume_1+volume_2+volume_3+volume_4)/5',  # 5日平均交易量
        '(high_0-low_0+high_1-low_1+high_2-low_2+high_3-low_3+high_4-low_4)/5',  # 5日平均振幅
        'market_cap_float_0',  # 流通市值
        'rank_market_cap_float_0',  # 流通市值排名
        'rank_pe_ttm_0',  # 市盈率TTM排名
        'mf_net_amount_5',  # 5日净主动买入额
        'avg_mf_net_amount_5',  # 5日平均净主动买入额
        'rank_avg_mf_net_amount_5',  # 5日平均净主动买入额排名
        'fs_net_profit_ttm_0',  # 归属母公司股东净利润TTM
        'fs_deducted_profit_0',  # 扣除经常性损益后净利润
        'fs_deducted_profit_ttm_0',  # 扣除经常性损益后净利润TTM
        'fs_net_profit_margin_ttm_0',  # 销售净利率TTM
        'rank_fs_operating_revenue_yoy_0',  # 营业收入同比增长率排名
        'fs_operating_revenue_qoq_0',  # 营业收入单季度环比增长率
        'rank_fs_operating_revenue_qoq_0',  # 营业收入单季度环比增长率排名
        'fs_eps_yoy_0',  # 每股收益同比增长率
        'fs_net_profit_yoy_0',  # 归属母公司股东净利润同比增长率
        'rank_fs_net_profit_yoy_0',  # 归属母公司股东净利润同比增长率排名
        'fs_net_profit_qoq_0',  # 归属母公司股东净利润单季度环比增长率
        'rank_fs_net_profit_qoq_0',  # 归属母公司股东净利润单季度环比增长率排名
        'fs_operating_revenue_yoy_0',  # 营业收入同比增长率
        'fs_roe_0',  # 净资产收益率
        'fs_roe_ttm_0',  # 净资产收益率TTM
        'fs_free_cash_flow_0',  # 企业自由现金流
        'fs_net_cash_flow_0',  # 经营活动产生的现金流量净额
        'sh_holder_avg_pct_0',  # 户均持股比例
        'ta_sma_5_0',  # 5日移动平均
        'ta_mom_10',  # 10日动量
        'ta_mfi_14_0',  # 14日资金流量指标
        'ta_atr_14_0',  # 14日平均真实波幅
        'ta_rsi_14_0',  # 14日相对强弱指标
        'ta_willr_14_0',  # 14日威廉指标
        'list_days_0',  # 已经上市的天数
        'list_board_0',  # 上市板
        'company_found_date_0',  # 公司成立天数
        'st_status_0',  # ST状态
    ]

# 给数据做标注:给每一行数据(样本)打分,一般分数越高表示越好
m1 = M.fast_auto_labeler.v5(
    instruments=conf.instruments, start_date=conf.start_date, end_date=conf.split_date,
    label_expr=conf.label_expr, hold_days=conf.hold_days,
    benchmark='000300.SHA', sell_at='open', buy_at='open')
# 计算特征数据
m2 = M.general_feature_extractor.v5(
    instruments=conf.instruments, start_date=conf.start_date, end_date=conf.split_date,
    features=conf.features)
# 数据预处理:缺失数据处理,数据规范化,T.get_stock_ranker_default_transforms为StockRanker模型做数据预处理
m3 = M.transform.v2(
    data=m2.data, transforms=T.get_stock_ranker_default_transforms(),
    drop_null=True, astype='int32', except_columns=['date', 'instrument'],
    clip_lower=0, clip_upper=200000000)
# 合并标注和特征数据
m4 = M.join.v2(data1=m1.data, data2=m3.data, on=['date', 'instrument'], sort=True)
# StockRanker机器学习训练
m5 = M.stock_ranker_train.v2(training_ds=m4.data, features=conf.features)


## 量化回测 https://bigquant.com/docs/module_trade.html
# 回测引擎:准备数据,只执行一次
def prepare(context):
    # context.start_date / end_date,回测的时候,为trader传入参数;在实盘运行的时候,由系统替换为实盘日期
    n1 = M.general_feature_extractor.v5(
        instruments=D.instruments(),
        start_date=context.start_date, end_date=context.end_date,
        model_id=context.options['model_id'])
    n2 = M.transform.v2(
        data=n1.data, transforms=T.get_stock_ranker_default_transforms(),
        drop_null=True, astype='int32', except_columns=['date', 'instrument'],
        clip_lower=0, clip_upper=200000000)
    n3 = M.stock_ranker_predict.v2(model_id=context.options['model_id'], data=n2.data)
    context.instruments = n3.instruments
    context.options['predictions'] = n3.predictions

# 回测引擎:初始化函数,只执行一次
def initialize(context):
    # 加载预测数据
    context.ranker_prediction = context.options['predictions'].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 = 5
    # 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

# 回测引擎:每日数据处理函数,每天执行一次
def handle_data(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')]

    # 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.options['hold_days'] # 是否在建仓期间(前 hold_days 天)
    cash_avg = context.portfolio.portfolio_value / context.options['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()}

    # 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)


# 调用交易引擎
m6 = M.trade.v1(
    instruments=None,
    start_date=conf.split_date,
    end_date=conf.end_date,
    prepare=prepare,
    initialize=initialize,
    handle_data=handle_data,
    order_price_field_buy='open',       # 表示 开盘 时买入
    order_price_field_sell='close',     # 表示 收盘 前卖出
    capital_base=1000000,               # 初始资金
    benchmark='000300.SHA',             # 比较基准,不影响回测结果
    # 通过 options 参数传递预测数据和参数给回测引擎
    options={'hold_days': conf.hold_days, 'model_id': m5.model_id}
)
[2017-07-21 13:43:58.049579] WARNING: bigquant: 此模块版本 M.fast_auto_labeler.v5 已不再维护,并可能在未来被删除:请更新到 fast_auto_labeler 最新版本
[2017-07-21 13:43:58.051188] INFO: bigquant: fast_auto_labeler.v5 start ..
[2017-07-21 13:43:58.055570] INFO: bigquant: hit cache
[2017-07-21 13:43:58.072102] INFO: bigquant: fast_auto_labeler.v5 end [0.020895s].
[2017-07-21 13:43:58.083890] INFO: bigquant: general_feature_extractor.v5 start ..
WARNING: unknown fields: {'ta_mom_10'}
[2017-07-21 13:44:30.706932] INFO: general_feature_extractor: year 2010, featurerows=431567
WARNING: unknown fields: {'ta_mom_10'}
[2017-07-21 13:45:07.652817] INFO: general_feature_extractor: year 2011, featurerows=511455
WARNING: unknown fields: {'ta_mom_10'}
[2017-07-21 13:45:48.001339] INFO: general_feature_extractor: year 2012, featurerows=565675
WARNING: unknown fields: {'ta_mom_10'}
[2017-07-21 13:46:32.261667] INFO: general_feature_extractor: year 2013, featurerows=564168
WARNING: unknown fields: {'ta_mom_10'}
[2017-07-21 13:47:22.327386] INFO: general_feature_extractor: year 2014, featurerows=569948
WARNING: unknown fields: {'ta_mom_10'}
[2017-07-21 13:47:50.643957] INFO: general_feature_extractor: year 2015, featurerows=0
[2017-07-21 13:47:50.653707] INFO: general_feature_extractor: total feature rows: 2642813
[2017-07-21 13:47:50.656156] INFO: bigquant: general_feature_extractor.v5 end [232.572257s].
[2017-07-21 13:47:50.670362] INFO: bigquant: transform.v2 start ..
[2017-07-21 13:47:58.969189] INFO: transform: transformed /y_2010, 147368/431567
[2017-07-21 13:48:09.089924] INFO: transform: transformed /y_2011, 196332/511455
[2017-07-21 13:48:20.747680] INFO: transform: transformed /y_2012, 533807/565675
[2017-07-21 13:48:36.290423] INFO: transform: transformed /y_2013, 544455/564168
[2017-07-21 13:48:54.143360] INFO: transform: transformed /y_2014, 548220/569948
[2017-07-21 13:48:54.172035] INFO: transform: transformed rows: 1970182/2642813
[2017-07-21 13:48:54.192770] INFO: bigquant: transform.v2 end [63.522399s].
[2017-07-21 13:48:54.203603] INFO: bigquant: join.v2 start ..
[2017-07-21 13:49:06.169588] INFO: join: /y_2010, rows=147275/147368, timetaken=2.969255s
[2017-07-21 13:49:10.381282] INFO: join: /y_2011, rows=196176/196332, timetaken=4.205215s
[2017-07-21 13:49:18.997423] INFO: join: /y_2012, rows=532945/533807, timetaken=8.604634s
[2017-07-21 13:49:27.853638] INFO: join: /y_2013, rows=543564/544455, timetaken=8.793034s
[2017-07-21 13:49:37.273682] INFO: join: /y_2014, rows=532174/548220, timetaken=9.365167s
[2017-07-21 13:49:37.408947] INFO: join: total result rows: 1952134
[2017-07-21 13:49:37.411184] INFO: bigquant: join.v2 end [43.207622s].
[2017-07-21 13:49:37.454882] WARNING: bigquant: 此模块版本 M.stock_ranker_train.v2 已不再维护,并可能在未来被删除:请更新到 stock_ranker_train 最新版本
[2017-07-21 13:49:37.455978] INFO: bigquant: stock_ranker_train.v2 start ..
[2017-07-21 13:49:50.208539] INFO: df2bin: prepare data: training ..
[2017-07-21 13:50:42.426306] INFO: stock_ranker_train: training: 1952134 rows
[2017-07-21 13:55:23.275561] INFO: bigquant: stock_ranker_train.v2 end [345.819551s].
[2017-07-21 13:55:23.319173] INFO: bigquant: backtest.v6 start ..
[2017-07-21 13:55:23.430400] INFO: bigquant: general_feature_extractor.v5 start ..
[2017-07-21 13:55:40.613247] INFO: general_feature_extractor: year 2015, featurerows=569698
[2017-07-21 13:56:06.597540] INFO: general_feature_extractor: year 2016, featurerows=641546
[2017-07-21 13:56:17.931828] INFO: general_feature_extractor: year 2017, featurerows=0
[2017-07-21 13:56:17.947423] INFO: general_feature_extractor: total feature rows: 1211244
[2017-07-21 13:56:17.949723] INFO: bigquant: general_feature_extractor.v5 end [54.519359s].
[2017-07-21 13:56:17.958572] INFO: bigquant: transform.v2 start ..
[2017-07-21 13:56:34.653151] INFO: transform: transformed /y_2015, 539183/569698
[2017-07-21 13:56:53.126535] INFO: transform: transformed /y_2016, 612123/641546
[2017-07-21 13:56:53.143184] INFO: transform: transformed rows: 1151306/1211244
[2017-07-21 13:56:53.165928] INFO: bigquant: transform.v2 end [35.207328s].
[2017-07-21 13:56:53.175790] INFO: bigquant: stock_ranker_predict.v2 start ..
[2017-07-21 13:56:58.209133] INFO: df2bin: prepare data: prediction ..
[2017-07-21 13:57:26.861626] INFO: stock_ranker_predict: prediction: 1151306 rows
[2017-07-21 13:58:43.535239] INFO: bigquant: stock_ranker_predict.v2 end [110.359401s].
[2017-07-21 13:59:09.467766] INFO: Performance: Simulated 488 trading days out of 488.
[2017-07-21 13:59:09.468951] INFO: Performance: first open: 2015-01-05 14:30:00+00:00
[2017-07-21 13:59:09.469795] INFO: Performance: last close: 2016-12-30 20:00:00+00:00
[注意] 有 3 笔卖出是在多天内完成的。当日卖出股票超过了当日股票交易的2.5%会出现这种情况。
  • 收益率354.13%
  • 年化收益率118.46%
  • 基准收益率-6.33%
  • 阿尔法1.2
  • 贝塔0.74
  • 夏普比率3.31
  • 收益波动率34.73%
  • 信息比率4.57
  • 最大回撤39.85%
[2017-07-21 13:59:11.346866] INFO: bigquant: backtest.v6 end [228.027678s].

:hamster:

(神龙斗士) #2

加油 :slightly_smiling_face: