复制链接

克隆策略

In [1]:

# 本代码由可视化策略环境自动生成 2023年1月14日 14:17
# 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。


# import empyrical

#计算函数，用到的列名也需要传入，例如close
#注意这里的df_total是整个数据集没有经过任何的groupby
def cal_fun(df_total, close, open, benchmark_close, benchmark_open):
    def get_label(df):
        #未来5日相对收益率
        relative_r = df.close.shift(-5) / df.open.shift(-1) - df.benchmark_close.shift(-5) / df.benchmark_open.shift(-1)
        #每天的相对收益率
        relative_r_daily = df.close.shift(-1) / df.close - df.benchmark_close.shift(-1) / df.benchmark_close
        #滚动求5日标准差
        std_ = relative_r_daily.rolling(5).apply(lambda x:np.std(x,ddof=1))
        #偏移成未来5日
        std1_ = std_.shift(-5)                                           
        df['label'] = relative_r / std1_
        return df
    
    df1=df_total.groupby(["instrument"]).apply(get_label)
    return df1['label']

    
m2_user_functions_bigquant_run = {
    'cal_fun':cal_fun
}

# 回测引擎：初始化函数，只执行一次
def m19_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 = 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
    context.options['hold_days'] = 5

# 回测引擎：每日数据处理函数，每天执行一次
def m19_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')]

    # 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.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，按机器学习算法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.portfolio.positions.items()}
        instruments = list(reversed(list(ranker_prediction.instrument[ranker_prediction.instrument.apply(
                lambda x: x in equities)])))

        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)])
    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 m19_prepare_bigquant_run(context):
    pass


m1 = M.instruments.v2(
    start_date='2018-01-01',
    end_date='2020-12-31',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0,
    m_cached=False
)

m2 = 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>`_

cal_fun(close, open, benchmark_close,benchmark_open)
# 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)
""",
    start_date='',
    end_date='',
    benchmark='000300.HIX',
    drop_na_label=True,
    cast_label_int=True,
    user_functions=m2_user_functions_bigquant_run
)

m3 = 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
"""
)

m15 = M.general_feature_extractor.v7(
    instruments=m1.data,
    features=m3.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m16 = M.derived_feature_extractor.v3(
    input_data=m15.data,
    features=m3.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False
)

m7 = M.join.v3(
    data1=m2.data,
    data2=m16.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m13 = M.dropnan.v1(
    input_data=m7.data
)

m6 = M.stock_ranker_train.v6(
    training_ds=m13.data,
    features=m3.data,
    learning_algorithm='排序',
    number_of_leaves=30,
    minimum_docs_per_leaf=1000,
    number_of_trees=20,
    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
)

m9 = M.instruments.v2(
    start_date=T.live_run_param('trading_date', '2021-01-01'),
    end_date=T.live_run_param('trading_date', '2021-12-31'),
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m17 = M.general_feature_extractor.v7(
    instruments=m9.data,
    features=m3.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m18 = M.derived_feature_extractor.v3(
    input_data=m17.data,
    features=m3.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False
)

m14 = M.dropnan.v1(
    input_data=m18.data
)

m8 = M.stock_ranker_predict.v5(
    model=m6.model,
    data=m14.data,
    m_lazy_run=False
)

m19 = M.trade.v4(
    instruments=m9.data,
    options_data=m8.predictions,
    start_date='',
    end_date='',
    initialize=m19_initialize_bigquant_run,
    handle_data=m19_handle_data_bigquant_run,
    prepare=m19_prepare_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='000300.HIX'
)

[2023-01-13 15:25:41.790202] INFO: moduleinvoker: instruments.v2 开始运行..

[2023-01-13 15:25:41.898228] INFO: moduleinvoker: instruments.v2 运行完成[0.108043s].

[2023-01-13 15:25:41.917773] INFO: moduleinvoker: advanced_auto_labeler.v2 开始运行..

[2023-01-13 15:25:45.513009] INFO: 自动标注(股票): 加载历史数据: 2647809 行

[2023-01-13 15:25:45.515323] INFO: 自动标注(股票): 开始标注 ..

[2023-01-13 15:30:18.599465] INFO: moduleinvoker: advanced_auto_labeler.v2 运行完成[276.681728s].

[2023-01-13 15:30:18.605410] INFO: moduleinvoker: input_features.v1 开始运行..

[2023-01-13 15:30:18.617814] INFO: moduleinvoker: 命中缓存

[2023-01-13 15:30:18.619475] INFO: moduleinvoker: input_features.v1 运行完成[0.014093s].

[2023-01-13 15:30:18.643216] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2023-01-13 15:30:19.793717] INFO: 基础特征抽取: 年份 2017, 特征行数=193398

[2023-01-13 15:30:22.956131] INFO: 基础特征抽取: 年份 2018, 特征行数=816987

[2023-01-13 15:30:26.217040] INFO: 基础特征抽取: 年份 2019, 特征行数=884867

[2023-01-13 15:30:29.354450] INFO: 基础特征抽取: 年份 2020, 特征行数=945961

[2023-01-13 15:30:29.653950] INFO: 基础特征抽取: 总行数: 2841213

[2023-01-13 15:30:29.773378] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[11.130164s].

[2023-01-13 15:30:29.845521] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2023-01-13 15:30:36.055120] INFO: derived_feature_extractor: 提取完成 avg_amount_0/avg_amount_5, 0.009s

[2023-01-13 15:30:36.065863] INFO: derived_feature_extractor: 提取完成 avg_amount_5/avg_amount_20, 0.009s

[2023-01-13 15:30:36.072907] INFO: derived_feature_extractor: 提取完成 rank_avg_amount_0/rank_avg_amount_5, 0.005s

[2023-01-13 15:30:36.079429] INFO: derived_feature_extractor: 提取完成 rank_avg_amount_5/rank_avg_amount_10, 0.005s

[2023-01-13 15:30:36.085802] INFO: derived_feature_extractor: 提取完成 rank_return_0/rank_return_5, 0.005s

[2023-01-13 15:30:36.094339] INFO: derived_feature_extractor: 提取完成 rank_return_5/rank_return_10, 0.006s

[2023-01-13 15:30:37.682013] INFO: derived_feature_extractor: /y_2017, 193398

[2023-01-13 15:30:39.402833] INFO: derived_feature_extractor: /y_2018, 816987

[2023-01-13 15:30:42.580886] INFO: derived_feature_extractor: /y_2019, 884867

[2023-01-13 15:30:45.585678] INFO: derived_feature_extractor: /y_2020, 945961

[2023-01-13 15:30:47.222256] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[17.376734s].

[2023-01-13 15:30:47.233653] INFO: moduleinvoker: join.v3 开始运行..

[2023-01-13 15:30:58.015777] INFO: join: /y_2017, 行数=0/193398, 耗时=1.207632s

[2023-01-13 15:31:01.246268] INFO: join: /y_2018, 行数=813500/816987, 耗时=3.227182s

[2023-01-13 15:31:04.819884] INFO: join: /y_2019, 行数=881275/884867, 耗时=3.563365s

[2023-01-13 15:31:08.782320] INFO: join: /y_2020, 行数=915235/945961, 耗时=3.953729s

[2023-01-13 15:31:08.990506] INFO: join: 最终行数: 2610010

[2023-01-13 15:31:09.020955] INFO: moduleinvoker: join.v3 运行完成[21.787301s].

[2023-01-13 15:31:09.033451] INFO: moduleinvoker: dropnan.v1 开始运行..

[2023-01-13 15:31:09.917248] INFO: dropnan: /y_2017, 0/0

[2023-01-13 15:31:13.487632] INFO: dropnan: /y_2018, 811820/813500

[2023-01-13 15:31:17.400175] INFO: dropnan: /y_2019, 877933/881275

[2023-01-13 15:31:21.570460] INFO: dropnan: /y_2020, 906957/915235

[2023-01-13 15:31:21.747384] INFO: dropnan: 行数: 2596710/2610010

[2023-01-13 15:31:21.763458] INFO: moduleinvoker: dropnan.v1 运行完成[12.730022s].

[2023-01-13 15:31:21.778420] INFO: moduleinvoker: stock_ranker_train.v6 开始运行..

[2023-01-13 15:31:33.239103] INFO: StockRanker: 特征预处理 ..

[2023-01-13 15:31:37.561790] INFO: StockRanker: prepare data: training ..

[2023-01-13 15:31:41.729986] INFO: StockRanker: sort ..

[2023-01-13 15:32:18.834398] INFO: StockRanker训练: 46fcfd14 准备训练: 2596710 行数

[2023-01-13 15:32:18.836739] INFO: StockRanker训练: AI模型训练，将在2596710*13=3375.72万数据上对模型训练进行20轮迭代训练。预计将需要11～21分钟。请耐心等待。

[2023-01-13 15:32:19.019690] INFO: StockRanker训练: 正在训练 ..

[2023-01-13 15:32:19.079115] INFO: StockRanker训练: 任务状态: Pending

[2023-01-13 15:32:29.181844] INFO: StockRanker训练: 任务状态: Running

[2023-01-13 15:33:49.574856] INFO: StockRanker训练: 00:01:17.1327549, finished iteration 1

[2023-01-13 15:33:59.622905] INFO: StockRanker训练: 00:01:29.9220534, finished iteration 2

[2023-01-13 15:34:09.671951] INFO: StockRanker训练: 00:01:44.4315032, finished iteration 3

[2023-01-13 15:34:29.776668] INFO: StockRanker训练: 00:01:57.4891718, finished iteration 4

[2023-01-13 15:34:39.823974] INFO: StockRanker训练: 00:02:10.7513531, finished iteration 5

[2023-01-13 15:34:49.868381] INFO: StockRanker训练: 00:02:23.7491569, finished iteration 6

[2023-01-13 15:35:09.959704] INFO: StockRanker训练: 00:02:37.4567513, finished iteration 7

[2023-01-13 15:35:20.009641] INFO: StockRanker训练: 00:02:50.8703353, finished iteration 8

[2023-01-13 15:35:30.055577] INFO: StockRanker训练: 00:03:04.7352961, finished iteration 9

[2023-01-13 15:35:50.156210] INFO: StockRanker训练: 00:03:18.6872625, finished iteration 10

[2023-01-13 15:36:00.201596] INFO: StockRanker训练: 00:03:32.9490879, finished iteration 11

[2023-01-13 15:36:20.292981] INFO: StockRanker训练: 00:03:49.9002122, finished iteration 12

[2023-01-13 15:36:40.387208] INFO: StockRanker训练: 00:04:07.1885246, finished iteration 13

[2023-01-13 15:36:50.430485] INFO: StockRanker训练: 00:04:23.9096019, finished iteration 14

[2023-01-13 15:37:10.521022] INFO: StockRanker训练: 00:04:40.6990201, finished iteration 15

[2023-01-13 15:37:30.651117] INFO: StockRanker训练: 00:04:57.2849706, finished iteration 16

[2023-01-13 15:37:40.829465] INFO: StockRanker训练: 00:05:14.1747979, finished iteration 17

[2023-01-13 15:38:00.927663] INFO: StockRanker训练: 00:05:31.1354489, finished iteration 18

[2023-01-13 15:38:21.013510] INFO: StockRanker训练: 00:05:48.4229480, finished iteration 19

[2023-01-13 15:38:31.061176] INFO: StockRanker训练: 00:06:05.8066670, finished iteration 20

[2023-01-13 15:38:31.063347] INFO: StockRanker训练: 任务状态: Succeeded

[2023-01-13 15:38:31.232376] INFO: moduleinvoker: stock_ranker_train.v6 运行完成[429.453945s].

[2023-01-13 15:38:31.239764] INFO: moduleinvoker: instruments.v2 开始运行..

[2023-01-13 15:38:31.248038] INFO: moduleinvoker: 命中缓存

[2023-01-13 15:38:31.250031] INFO: moduleinvoker: instruments.v2 运行完成[0.010273s].

[2023-01-13 15:38:31.266270] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2023-01-13 15:38:31.277361] INFO: moduleinvoker: 命中缓存

[2023-01-13 15:38:31.279351] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[0.013108s].

[2023-01-13 15:38:31.287584] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2023-01-13 15:38:31.301941] INFO: moduleinvoker: 命中缓存

[2023-01-13 15:38:31.304148] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[0.016579s].

[2023-01-13 15:38:31.314703] INFO: moduleinvoker: dropnan.v1 开始运行..

[2023-01-13 15:38:31.325847] INFO: moduleinvoker: 命中缓存

[2023-01-13 15:38:31.327728] INFO: moduleinvoker: dropnan.v1 运行完成[0.013032s].

[2023-01-13 15:38:31.341451] INFO: moduleinvoker: stock_ranker_predict.v5 开始运行..

[2023-01-13 15:38:32.254341] INFO: StockRanker预测: /y_2020 ..

[2023-01-13 15:38:34.651232] INFO: StockRanker预测: /y_2021 ..

[2023-01-13 15:38:38.622796] INFO: moduleinvoker: stock_ranker_predict.v5 运行完成[7.28134s].

[2023-01-13 15:38:42.510698] INFO: moduleinvoker: backtest.v8 开始运行..

[2023-01-13 15:38:42.518988] INFO: backtest: biglearning backtest:V8.6.3

[2023-01-13 15:38:42.521379] INFO: backtest: product_type:stock by specified

[2023-01-13 15:38:42.587447] INFO: moduleinvoker: cached.v2 开始运行..

[2023-01-13 15:38:42.599660] INFO: moduleinvoker: 命中缓存

[2023-01-13 15:38:42.601553] INFO: moduleinvoker: cached.v2 运行完成[0.014135s].

[2023-01-13 15:38:53.785010] INFO: backtest: algo history_data=DataSource(a929ac626e724fc5a53ed51cf775e582T)

[2023-01-13 15:38:53.787357] INFO: algo: TradingAlgorithm V1.8.9

[2023-01-13 15:38:56.564367] INFO: algo: trading transform...

[2023-01-13 15:39:05.487764] INFO: Performance: Simulated 243 trading days out of 243.

[2023-01-13 15:39:05.489580] INFO: Performance: first open: 2021-01-04 09:30:00+00:00

[2023-01-13 15:39:05.491488] INFO: Performance: last close: 2021-12-31 15:00:00+00:00

[2023-01-13 15:39:10.903798] INFO: moduleinvoker: backtest.v8 运行完成[28.393112s].

[2023-01-13 15:39:10.905686] INFO: moduleinvoker: trade.v4 运行完成[32.262227s].

设置评估测试数据集，查看训练曲线
[视频教程]StockRanker训练曲线

bigcharts-data-start/{"__type":"tabs","__id":"bigchart-3177bc8054994b9eb8912feb004e9d77"}/bigcharts-data-end

收益率66.31%
年化收益率69.47%
基准收益率-5.2%
阿尔法0.74
贝塔0.47
夏普比率2.29
胜率0.54
盈亏比1.28
收益波动率22.93%
信息比率0.16
最大回撤9.96%

bigcharts-data-start/{"__type":"tabs","__id":"bigchart-8f1ba6a575804b5785f6911b6c6ccc89"}/bigcharts-data-end