克隆策略

Transformer在量化选股中的应用研究¶

Transformer:Attention is all you need¶

The naive transformer implemented here for financial time series prediction follows the paper "Attention is all you need":

Given the input (N, T, F),

An embedding layer that maps the input (N, T, F) to representation (N, T, F’);
A positional encoding layer that adds the positional sigmoid;
An encoder that consists of several encoding layers, each of which uses a self-attention layer as the computing module (function of query, key, and value).
A decoder that consists of an MLP (or a Linear layer) that maps the representation of the last time (N, 1, F') into output (N, 1).

In [1]:

# 本代码由可视化策略环境自动生成 2021年10月26日 16:27
# 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。


# 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 m33_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    from sklearn.model_selection import train_test_split
    # train data
    train_data = input_1.read()
    x_train, x_val, y_train, y_val = train_test_split(train_data["x"], train_data['y'], shuffle=True, random_state=2021)
    # val data
    test_data = input_2.read()
    x_test = test_data["x"]
    
    from bigmodels.models.transformer import Transformer
    
    model = Transformer(input_dim=98, embed_dim=256, nhead=8, num_layers=6, dropout=0.1)
    model.compile(device="cuda:0")
    model.fit(x_train, y_train, val_data=(x_val, y_val), batch_size=2048, epochs=10, verbose=1, num_workers=2)
    
    # model.fit(train_data["x"], train_data['y'], batch_size=1024, epochs=2, verbose=1, num_workers=2)
    output = model.predict(x_test)
    
    data_1 = DataSource.write_pickle(output)
    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 m33_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 m41_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    pred_label = input_1.read_pickle()
    
    df = input_2.read_df()
    df = pd.DataFrame({'pred_label':pred_label[:], 'instrument':df.instrument, 'date':df.date})
    df.sort_values(['date','pred_label'],inplace=True, ascending=[True,False])
    return Outputs(data_1=DataSource.write_df(df), 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 m41_post_run_bigquant_run(outputs):
    return outputs

# 回测引擎：初始化函数，只执行一次
def m42_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=0.001, sell_cost=0.001, 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 = 20
    # 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 m42_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.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.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 m42_prepare_bigquant_run(context):
    pass


m22 = M.instruments.v2(
    start_date='2010-01-01',
    end_date='2017-12-31',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m23 = M.advanced_auto_labeler.v2(
    instruments=m22.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/data_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/big_expr.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, -5) / 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))

# 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.SHA',
    drop_na_label=True,
    cast_label_int=False
)

m1 = M.standardlize.v8(
    input_1=m23.data,
    columns_input='label'
)

m24 = M.input_features.v1(
    features="""close_0
open_0
high_0
low_0 
amount_0
turn_0 
return_0
 
close_1
open_1
high_1
low_1
return_1
amount_1
turn_1
 
close_2
open_2
high_2
low_2
amount_2
turn_2
return_2
 
close_3
open_3
high_3
low_3
amount_3
turn_3
return_3
 
close_4
open_4
high_4
low_4
amount_4
turn_4
return_4
 
mean(close_0, 5)
mean(low_0, 5)
mean(open_0, 5)
mean(high_0, 5)
mean(turn_0, 5)
mean(amount_0, 5)
mean(return_0, 5)
 
ts_max(close_0, 5)
ts_max(low_0, 5)
ts_max(open_0, 5)
ts_max(high_0, 5)
ts_max(turn_0, 5)
ts_max(amount_0, 5)
ts_max(return_0, 5)
 
ts_min(close_0, 5)
ts_min(low_0, 5)
ts_min(open_0, 5)
ts_min(high_0, 5)
ts_min(turn_0, 5)
ts_min(amount_0, 5)
ts_min(return_0, 5) 
 
std(close_0, 5)
std(low_0, 5)
std(open_0, 5)
std(high_0, 5)
std(turn_0, 5)
std(amount_0, 5)
std(return_0, 5)
 
ts_rank(close_0, 5)
ts_rank(low_0, 5)
ts_rank(open_0, 5)
ts_rank(high_0, 5)
ts_rank(turn_0, 5)
ts_rank(amount_0, 5)
ts_rank(return_0, 5)
 
decay_linear(close_0, 5)
decay_linear(low_0, 5)
decay_linear(open_0, 5)
decay_linear(high_0, 5)
decay_linear(turn_0, 5)
decay_linear(amount_0, 5)
decay_linear(return_0, 5)
 
correlation(volume_0, return_0, 5)
correlation(volume_0, high_0, 5)
correlation(volume_0, low_0, 5)
correlation(volume_0, close_0, 5)
correlation(volume_0, open_0, 5)
correlation(volume_0, turn_0, 5)
  
correlation(return_0, high_0, 5)
correlation(return_0, low_0, 5)
correlation(return_0, close_0, 5)
correlation(return_0, open_0, 5)
correlation(return_0, turn_0, 5)
 
correlation(high_0, low_0, 5)
correlation(high_0, close_0, 5)
correlation(high_0, open_0, 5)
correlation(high_0, turn_0, 5)
 
correlation(low_0, close_0, 5)
correlation(low_0, open_0, 5)
correlation(low_0, turn_0, 5)
 
correlation(close_0, open_0, 5)
correlation(close_0, turn_0, 5)

correlation(open_0, turn_0, 5)"""
)

m27 = M.general_feature_extractor.v7(
    instruments=m22.data,
    features=m24.data,
    start_date='',
    end_date='',
    before_start_days=10
)

m28 = M.derived_feature_extractor.v3(
    input_data=m27.data,
    features=m24.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=True,
    remove_extra_columns=False
)

m34 = M.standardlize.v8(
    input_1=m28.data,
    input_2=m24.data,
    columns_input='[]'
)

m35 = M.fillnan.v1(
    input_data=m34.data,
    features=m24.data,
    fill_value='0.0'
)

m25 = M.join.v3(
    data1=m1.data,
    data2=m35.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m2 = M.dl_convert_to_bin.v2(
    input_data=m25.data,
    features=m24.data,
    window_size=5,
    feature_clip=3,
    flatten=False,
    window_along_col='instrument'
)

m26 = M.instruments.v2(
    start_date='2018-01-01',
    end_date='2021-07-01',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m29 = M.general_feature_extractor.v7(
    instruments=m26.data,
    features=m24.data,
    start_date='',
    end_date='',
    before_start_days=10
)

m30 = M.derived_feature_extractor.v3(
    input_data=m29.data,
    features=m24.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=True,
    remove_extra_columns=False
)

m37 = M.standardlize.v8(
    input_1=m30.data,
    input_2=m24.data,
    columns_input='[]'
)

m36 = M.fillnan.v1(
    input_data=m37.data,
    features=m24.data,
    fill_value='0.0'
)

m32 = M.dl_convert_to_bin.v2(
    input_data=m36.data,
    features=m24.data,
    window_size=5,
    feature_clip=3,
    flatten=False,
    window_along_col='instrument'
)

m33 = M.cached.v3(
    input_1=m2.data,
    input_2=m32.data,
    run=m33_run_bigquant_run,
    post_run=m33_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports='',
    m_cached=False
)

m41 = M.cached.v3(
    input_1=m33.data_1,
    input_2=m30.data,
    run=m41_run_bigquant_run,
    post_run=m41_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports='',
    m_cached=False
)

m42 = M.trade.v4(
    instruments=m26.data,
    options_data=m41.data_1,
    start_date='',
    end_date='',
    initialize=m42_initialize_bigquant_run,
    handle_data=m42_handle_data_bigquant_run,
    prepare=m42_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='后复权',
    product_type='股票',
    plot_charts=True,
    backtest_only=False,
    benchmark='000300.SHA'
)

[2021-10-26 11:53:50.153481] INFO: moduleinvoker: instruments.v2 开始运行..

[2021-10-26 11:53:50.171521] INFO: moduleinvoker: 命中缓存

[2021-10-26 11:53:50.173299] INFO: moduleinvoker: instruments.v2 运行完成[0.019842s].

[2021-10-26 11:53:50.186849] INFO: moduleinvoker: advanced_auto_labeler.v2 开始运行..

[2021-10-26 11:53:54.220897] INFO: 自动标注(股票): 加载历史数据: 4597290 行

[2021-10-26 11:53:54.222802] INFO: 自动标注(股票): 开始标注 ..

[2021-10-26 11:54:00.037485] INFO: moduleinvoker: advanced_auto_labeler.v2 运行完成[9.850633s].

[2021-10-26 11:54:00.047115] INFO: moduleinvoker: standardlize.v8 开始运行..

[2021-10-26 11:54:35.140195] INFO: moduleinvoker: standardlize.v8 运行完成[35.093074s].

[2021-10-26 11:54:35.149912] INFO: moduleinvoker: input_features.v1 开始运行..

[2021-10-26 11:54:35.157407] INFO: moduleinvoker: 命中缓存

[2021-10-26 11:54:35.158889] INFO: moduleinvoker: input_features.v1 运行完成[0.008981s].

[2021-10-26 11:54:35.180277] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2021-10-26 11:54:35.852280] INFO: 基础特征抽取: 年份 2009, 特征行数=12795

[2021-10-26 11:54:37.641108] INFO: 基础特征抽取: 年份 2010, 特征行数=431567

[2021-10-26 11:54:40.014269] INFO: 基础特征抽取: 年份 2011, 特征行数=511455

[2021-10-26 11:54:42.560162] INFO: 基础特征抽取: 年份 2012, 特征行数=565675

[2021-10-26 11:54:45.498635] INFO: 基础特征抽取: 年份 2013, 特征行数=564168

[2021-10-26 11:54:48.333216] INFO: 基础特征抽取: 年份 2014, 特征行数=569948

[2021-10-26 11:54:51.292088] INFO: 基础特征抽取: 年份 2015, 特征行数=569698

[2021-10-26 11:54:54.306600] INFO: 基础特征抽取: 年份 2016, 特征行数=641546

[2021-10-26 11:54:57.600060] INFO: 基础特征抽取: 年份 2017, 特征行数=743233

[2021-10-26 11:54:57.785779] INFO: 基础特征抽取: 总行数: 4610085

[2021-10-26 11:54:57.798430] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[22.618144s].

[2021-10-26 11:54:57.813805] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2021-10-26 11:55:12.355656] INFO: derived_feature_extractor: 提取完成 mean(close_0, 5), 3.105s

[2021-10-26 11:55:15.505062] INFO: derived_feature_extractor: 提取完成 mean(low_0, 5), 3.148s

[2021-10-26 11:55:18.670265] INFO: derived_feature_extractor: 提取完成 mean(open_0, 5), 3.163s

[2021-10-26 11:55:22.073671] INFO: derived_feature_extractor: 提取完成 mean(high_0, 5), 3.402s

[2021-10-26 11:55:25.606597] INFO: derived_feature_extractor: 提取完成 mean(turn_0, 5), 3.531s

[2021-10-26 11:55:29.168257] INFO: derived_feature_extractor: 提取完成 mean(amount_0, 5), 3.559s

[2021-10-26 11:55:32.639465] INFO: derived_feature_extractor: 提取完成 mean(return_0, 5), 3.470s

[2021-10-26 11:55:36.130133] INFO: derived_feature_extractor: 提取完成 ts_max(close_0, 5), 3.489s

[2021-10-26 11:55:39.607039] INFO: derived_feature_extractor: 提取完成 ts_max(low_0, 5), 3.475s

[2021-10-26 11:55:43.033029] INFO: derived_feature_extractor: 提取完成 ts_max(open_0, 5), 3.424s

[2021-10-26 11:55:46.471137] INFO: derived_feature_extractor: 提取完成 ts_max(high_0, 5), 3.436s

[2021-10-26 11:55:50.220594] INFO: derived_feature_extractor: 提取完成 ts_max(turn_0, 5), 3.748s

[2021-10-26 11:55:53.811971] INFO: derived_feature_extractor: 提取完成 ts_max(amount_0, 5), 3.590s

[2021-10-26 11:55:57.420308] INFO: derived_feature_extractor: 提取完成 ts_max(return_0, 5), 3.606s

[2021-10-26 11:56:00.844461] INFO: derived_feature_extractor: 提取完成 ts_min(close_0, 5), 3.422s

[2021-10-26 11:56:04.302785] INFO: derived_feature_extractor: 提取完成 ts_min(low_0, 5), 3.456s

[2021-10-26 11:56:07.776412] INFO: derived_feature_extractor: 提取完成 ts_min(open_0, 5), 3.471s

[2021-10-26 11:56:11.216635] INFO: derived_feature_extractor: 提取完成 ts_min(high_0, 5), 3.439s

[2021-10-26 11:56:14.652846] INFO: derived_feature_extractor: 提取完成 ts_min(turn_0, 5), 3.435s

[2021-10-26 11:56:20.422620] INFO: derived_feature_extractor: 提取完成 ts_min(amount_0, 5), 5.768s

[2021-10-26 11:56:23.895356] INFO: derived_feature_extractor: 提取完成 ts_min(return_0, 5), 3.471s

[2021-10-26 11:56:27.562111] INFO: derived_feature_extractor: 提取完成 std(close_0, 5), 3.665s

[2021-10-26 11:56:31.071210] INFO: derived_feature_extractor: 提取完成 std(low_0, 5), 3.507s

[2021-10-26 11:56:34.599965] INFO: derived_feature_extractor: 提取完成 std(open_0, 5), 3.527s

[2021-10-26 11:56:38.156569] INFO: derived_feature_extractor: 提取完成 std(high_0, 5), 3.554s

[2021-10-26 11:56:41.780756] INFO: derived_feature_extractor: 提取完成 std(turn_0, 5), 3.623s

[2021-10-26 11:56:45.286708] INFO: derived_feature_extractor: 提取完成 std(amount_0, 5), 3.504s

[2021-10-26 11:56:48.737082] INFO: derived_feature_extractor: 提取完成 std(return_0, 5), 3.449s

[2021-10-26 11:57:03.732952] INFO: derived_feature_extractor: 提取完成 ts_rank(close_0, 5), 14.994s

[2021-10-26 11:57:18.459438] INFO: derived_feature_extractor: 提取完成 ts_rank(low_0, 5), 14.725s

[2021-10-26 11:57:33.122956] INFO: derived_feature_extractor: 提取完成 ts_rank(open_0, 5), 14.662s

[2021-10-26 11:57:47.786432] INFO: derived_feature_extractor: 提取完成 ts_rank(high_0, 5), 14.662s

[2021-10-26 11:58:02.666538] INFO: derived_feature_extractor: 提取完成 ts_rank(turn_0, 5), 14.878s

[2021-10-26 11:58:18.017712] INFO: derived_feature_extractor: 提取完成 ts_rank(amount_0, 5), 15.350s

[2021-10-26 11:58:31.295005] INFO: derived_feature_extractor: 提取完成 ts_rank(return_0, 5), 13.276s

[2021-10-26 11:58:40.218690] INFO: derived_feature_extractor: 提取完成 decay_linear(close_0, 5), 8.922s

[2021-10-26 11:58:49.214164] INFO: derived_feature_extractor: 提取完成 decay_linear(low_0, 5), 8.993s

[2021-10-26 11:58:59.132243] INFO: derived_feature_extractor: 提取完成 decay_linear(open_0, 5), 9.917s

[2021-10-26 11:59:08.514253] INFO: derived_feature_extractor: 提取完成 decay_linear(high_0, 5), 9.380s

[2021-10-26 11:59:18.789320] INFO: derived_feature_extractor: 提取完成 decay_linear(turn_0, 5), 10.273s

[2021-10-26 11:59:28.503915] INFO: derived_feature_extractor: 提取完成 decay_linear(amount_0, 5), 9.713s

[2021-10-26 11:59:37.544933] INFO: derived_feature_extractor: 提取完成 decay_linear(return_0, 5), 9.039s

[2021-10-26 12:00:02.975498] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, return_0, 5), 25.429s

[2021-10-26 12:00:27.372198] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, high_0, 5), 24.395s

[2021-10-26 12:00:52.934548] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, low_0, 5), 25.561s

[2021-10-26 12:01:20.010069] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, close_0, 5), 27.074s

[2021-10-26 12:01:44.298519] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, open_0, 5), 24.287s

[2021-10-26 12:02:08.581897] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, turn_0, 5), 24.281s

[2021-10-26 12:02:34.001895] INFO: derived_feature_extractor: 提取完成 correlation(return_0, high_0, 5), 25.418s

[2021-10-26 12:02:59.017000] INFO: derived_feature_extractor: 提取完成 correlation(return_0, low_0, 5), 25.012s

[2021-10-26 12:03:22.921364] INFO: derived_feature_extractor: 提取完成 correlation(return_0, close_0, 5), 23.902s

[2021-10-26 12:03:48.729838] INFO: derived_feature_extractor: 提取完成 correlation(return_0, open_0, 5), 25.807s

[2021-10-26 12:04:15.353114] INFO: derived_feature_extractor: 提取完成 correlation(return_0, turn_0, 5), 26.622s

[2021-10-26 12:04:41.273675] INFO: derived_feature_extractor: 提取完成 correlation(high_0, low_0, 5), 25.919s

[2021-10-26 12:05:06.007658] INFO: derived_feature_extractor: 提取完成 correlation(high_0, close_0, 5), 24.732s

[2021-10-26 12:05:33.035505] INFO: derived_feature_extractor: 提取完成 correlation(high_0, open_0, 5), 27.026s

[2021-10-26 12:05:58.033689] INFO: derived_feature_extractor: 提取完成 correlation(high_0, turn_0, 5), 24.996s

[2021-10-26 12:06:24.976763] INFO: derived_feature_extractor: 提取完成 correlation(low_0, close_0, 5), 26.941s

[2021-10-26 12:06:48.961616] INFO: derived_feature_extractor: 提取完成 correlation(low_0, open_0, 5), 23.983s

[2021-10-26 12:07:13.420652] INFO: derived_feature_extractor: 提取完成 correlation(low_0, turn_0, 5), 24.457s

[2021-10-26 12:07:37.684751] INFO: derived_feature_extractor: 提取完成 correlation(close_0, open_0, 5), 24.262s

[2021-10-26 12:08:01.753210] INFO: derived_feature_extractor: 提取完成 correlation(close_0, turn_0, 5), 24.067s

[2021-10-26 12:08:25.407178] INFO: derived_feature_extractor: 提取完成 correlation(open_0, turn_0, 5), 23.652s

[2021-10-26 12:08:36.650192] INFO: derived_feature_extractor: /y_2009, 12795

[2021-10-26 12:08:37.734691] INFO: derived_feature_extractor: /y_2010, 431567

[2021-10-26 12:08:41.034377] INFO: derived_feature_extractor: /y_2011, 511455

[2021-10-26 12:08:44.912275] INFO: derived_feature_extractor: /y_2012, 565675

[2021-10-26 12:08:55.839108] INFO: derived_feature_extractor: /y_2013, 564168

[2021-10-26 12:09:01.819675] INFO: derived_feature_extractor: /y_2014, 569948

[2021-10-26 12:09:06.178728] INFO: derived_feature_extractor: /y_2015, 569698

[2021-10-26 12:09:10.828357] INFO: derived_feature_extractor: /y_2016, 641546

[2021-10-26 12:09:15.870005] INFO: derived_feature_extractor: /y_2017, 743233

[2021-10-26 12:09:19.815031] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[862.001215s].

[2021-10-26 12:09:19.821489] INFO: moduleinvoker: standardlize.v8 开始运行..

[2021-10-26 12:18:12.229323] INFO: moduleinvoker: standardlize.v8 运行完成[532.407828s].

[2021-10-26 12:18:12.244742] INFO: moduleinvoker: fillnan.v1 开始运行..

[2021-10-26 12:18:48.164191] INFO: moduleinvoker: fillnan.v1 运行完成[35.919437s].

[2021-10-26 12:18:48.180423] INFO: moduleinvoker: join.v3 开始运行..

[2021-10-26 12:20:17.262404] INFO: join: /data, 行数=4551595/4593884, 耗时=82.428297s

[2021-10-26 12:20:17.585025] INFO: join: 最终行数: 4551595

[2021-10-26 12:20:17.611010] INFO: moduleinvoker: join.v3 运行完成[89.430583s].

[2021-10-26 12:20:17.635495] INFO: moduleinvoker: dl_convert_to_bin.v2 开始运行..

[2021-10-26 12:27:25.442033] INFO: moduleinvoker: dl_convert_to_bin.v2 运行完成[427.806539s].

[2021-10-26 12:27:25.449393] INFO: moduleinvoker: instruments.v2 开始运行..

[2021-10-26 12:27:25.582275] INFO: moduleinvoker: instruments.v2 运行完成[0.132887s].

[2021-10-26 12:27:25.594975] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2021-10-26 12:27:26.711155] INFO: 基础特征抽取: 年份 2017, 特征行数=19477

[2021-10-26 12:27:30.429043] INFO: 基础特征抽取: 年份 2018, 特征行数=816987

[2021-10-26 12:27:34.965009] INFO: 基础特征抽取: 年份 2019, 特征行数=884867

[2021-10-26 12:27:39.441845] INFO: 基础特征抽取: 年份 2020, 特征行数=945961

[2021-10-26 12:27:42.652818] INFO: 基础特征抽取: 年份 2021, 特征行数=503096

[2021-10-26 12:27:42.719656] INFO: 基础特征抽取: 总行数: 3170388

[2021-10-26 12:27:42.726163] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[17.131184s].

[2021-10-26 12:27:42.733879] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2021-10-26 12:27:52.363850] INFO: derived_feature_extractor: 提取完成 mean(close_0, 5), 1.940s

[2021-10-26 12:27:54.283908] INFO: derived_feature_extractor: 提取完成 mean(low_0, 5), 1.918s

[2021-10-26 12:27:56.255265] INFO: derived_feature_extractor: 提取完成 mean(open_0, 5), 1.970s

[2021-10-26 12:27:58.139536] INFO: derived_feature_extractor: 提取完成 mean(high_0, 5), 1.882s

[2021-10-26 12:28:00.141823] INFO: derived_feature_extractor: 提取完成 mean(turn_0, 5), 2.001s

[2021-10-26 12:28:02.199664] INFO: derived_feature_extractor: 提取完成 mean(amount_0, 5), 2.056s

[2021-10-26 12:28:04.246438] INFO: derived_feature_extractor: 提取完成 mean(return_0, 5), 2.045s

[2021-10-26 12:28:06.392845] INFO: derived_feature_extractor: 提取完成 ts_max(close_0, 5), 2.145s

[2021-10-26 12:28:08.543476] INFO: derived_feature_extractor: 提取完成 ts_max(low_0, 5), 2.149s

[2021-10-26 12:28:10.632842] INFO: derived_feature_extractor: 提取完成 ts_max(open_0, 5), 2.088s

[2021-10-26 12:28:12.668501] INFO: derived_feature_extractor: 提取完成 ts_max(high_0, 5), 2.034s

[2021-10-26 12:28:14.639460] INFO: derived_feature_extractor: 提取完成 ts_max(turn_0, 5), 1.969s

[2021-10-26 12:28:16.574115] INFO: derived_feature_extractor: 提取完成 ts_max(amount_0, 5), 1.933s

[2021-10-26 12:28:18.513672] INFO: derived_feature_extractor: 提取完成 ts_max(return_0, 5), 1.938s

[2021-10-26 12:28:20.452093] INFO: derived_feature_extractor: 提取完成 ts_min(close_0, 5), 1.937s

[2021-10-26 12:28:22.441648] INFO: derived_feature_extractor: 提取完成 ts_min(low_0, 5), 1.988s

[2021-10-26 12:28:24.446836] INFO: derived_feature_extractor: 提取完成 ts_min(open_0, 5), 2.003s

[2021-10-26 12:28:26.413216] INFO: derived_feature_extractor: 提取完成 ts_min(high_0, 5), 1.965s

[2021-10-26 12:28:28.444456] INFO: derived_feature_extractor: 提取完成 ts_min(turn_0, 5), 2.030s

[2021-10-26 12:28:30.448992] INFO: derived_feature_extractor: 提取完成 ts_min(amount_0, 5), 2.003s

[2021-10-26 12:28:32.464012] INFO: derived_feature_extractor: 提取完成 ts_min(return_0, 5), 2.013s

[2021-10-26 12:28:34.498931] INFO: derived_feature_extractor: 提取完成 std(close_0, 5), 2.033s

[2021-10-26 12:28:36.604983] INFO: derived_feature_extractor: 提取完成 std(low_0, 5), 2.104s

[2021-10-26 12:28:38.689591] INFO: derived_feature_extractor: 提取完成 std(open_0, 5), 2.083s

[2021-10-26 12:28:40.769011] INFO: derived_feature_extractor: 提取完成 std(high_0, 5), 2.078s

[2021-10-26 12:28:42.794530] INFO: derived_feature_extractor: 提取完成 std(turn_0, 5), 2.024s

[2021-10-26 12:28:44.875730] INFO: derived_feature_extractor: 提取完成 std(amount_0, 5), 2.080s

[2021-10-26 12:28:46.898182] INFO: derived_feature_extractor: 提取完成 std(return_0, 5), 2.021s

[2021-10-26 12:28:56.208391] INFO: derived_feature_extractor: 提取完成 ts_rank(close_0, 5), 9.309s

[2021-10-26 12:29:06.116166] INFO: derived_feature_extractor: 提取完成 ts_rank(low_0, 5), 9.906s

[2021-10-26 12:29:14.331373] INFO: derived_feature_extractor: 提取完成 ts_rank(open_0, 5), 8.214s

[2021-10-26 12:29:22.487410] INFO: derived_feature_extractor: 提取完成 ts_rank(high_0, 5), 8.155s

[2021-10-26 12:29:31.381519] INFO: derived_feature_extractor: 提取完成 ts_rank(turn_0, 5), 8.893s

[2021-10-26 12:29:40.152395] INFO: derived_feature_extractor: 提取完成 ts_rank(amount_0, 5), 8.769s

[2021-10-26 12:29:49.191599] INFO: derived_feature_extractor: 提取完成 ts_rank(return_0, 5), 9.037s

[2021-10-26 12:29:54.872129] INFO: derived_feature_extractor: 提取完成 decay_linear(close_0, 5), 5.679s

[2021-10-26 12:30:00.766686] INFO: derived_feature_extractor: 提取完成 decay_linear(low_0, 5), 5.893s

[2021-10-26 12:30:06.451341] INFO: derived_feature_extractor: 提取完成 decay_linear(open_0, 5), 5.683s

[2021-10-26 12:30:12.246862] INFO: derived_feature_extractor: 提取完成 decay_linear(high_0, 5), 5.794s

[2021-10-26 12:30:17.725530] INFO: derived_feature_extractor: 提取完成 decay_linear(turn_0, 5), 5.477s

[2021-10-26 12:30:23.244134] INFO: derived_feature_extractor: 提取完成 decay_linear(amount_0, 5), 5.517s

[2021-10-26 12:30:28.655429] INFO: derived_feature_extractor: 提取完成 decay_linear(return_0, 5), 5.410s

[2021-10-26 12:30:56.206949] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, return_0, 5), 27.550s

[2021-10-26 12:31:23.410111] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, high_0, 5), 27.202s

[2021-10-26 12:31:50.606612] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, low_0, 5), 27.195s

[2021-10-26 12:32:17.829087] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, close_0, 5), 27.221s

[2021-10-26 12:32:46.023033] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, open_0, 5), 28.192s

[2021-10-26 12:33:13.672546] INFO: derived_feature_extractor: 提取完成 correlation(volume_0, turn_0, 5), 27.648s

[2021-10-26 12:33:42.205336] INFO: derived_feature_extractor: 提取完成 correlation(return_0, high_0, 5), 28.531s

[2021-10-26 12:34:11.623303] INFO: derived_feature_extractor: 提取完成 correlation(return_0, low_0, 5), 29.416s

[2021-10-26 12:34:39.994834] INFO: derived_feature_extractor: 提取完成 correlation(return_0, close_0, 5), 28.370s

[2021-10-26 12:35:08.699852] INFO: derived_feature_extractor: 提取完成 correlation(return_0, open_0, 5), 28.703s

[2021-10-26 12:35:37.518479] INFO: derived_feature_extractor: 提取完成 correlation(return_0, turn_0, 5), 28.817s

[2021-10-26 12:36:05.670344] INFO: derived_feature_extractor: 提取完成 correlation(high_0, low_0, 5), 28.150s

[2021-10-26 12:36:33.215872] INFO: derived_feature_extractor: 提取完成 correlation(high_0, close_0, 5), 27.544s

[2021-10-26 12:37:01.105531] INFO: derived_feature_extractor: 提取完成 correlation(high_0, open_0, 5), 27.888s

[2021-10-26 12:37:29.016628] INFO: derived_feature_extractor: 提取完成 correlation(high_0, turn_0, 5), 27.909s

[2021-10-26 12:37:56.723779] INFO: derived_feature_extractor: 提取完成 correlation(low_0, close_0, 5), 27.705s

[2021-10-26 12:38:25.542625] INFO: derived_feature_extractor: 提取完成 correlation(low_0, open_0, 5), 28.817s

[2021-10-26 12:38:55.289339] INFO: derived_feature_extractor: 提取完成 correlation(low_0, turn_0, 5), 29.745s

[2021-10-26 12:39:24.894432] INFO: derived_feature_extractor: 提取完成 correlation(close_0, open_0, 5), 29.603s

[2021-10-26 12:39:52.029630] INFO: derived_feature_extractor: 提取完成 correlation(close_0, turn_0, 5), 27.133s

[2021-10-26 12:40:19.871997] INFO: derived_feature_extractor: 提取完成 correlation(open_0, turn_0, 5), 27.841s

[2021-10-26 12:40:22.726691] INFO: derived_feature_extractor: /y_2017, 19477

[2021-10-26 12:40:24.561747] INFO: derived_feature_extractor: /y_2018, 816987

[2021-10-26 12:40:29.724688] INFO: derived_feature_extractor: /y_2019, 884867

[2021-10-26 12:40:35.368188] INFO: derived_feature_extractor: /y_2020, 945961

[2021-10-26 12:40:40.349933] INFO: derived_feature_extractor: /y_2021, 503096

[2021-10-26 12:40:42.644037] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[779.910146s].

[2021-10-26 12:40:42.651622] INFO: moduleinvoker: standardlize.v8 开始运行..

[2021-10-26 12:45:48.480882] INFO: moduleinvoker: standardlize.v8 运行完成[305.829258s].

[2021-10-26 12:45:48.489918] INFO: moduleinvoker: fillnan.v1 开始运行..

[2021-10-26 12:46:13.162504] INFO: moduleinvoker: fillnan.v1 运行完成[24.672576s].

[2021-10-26 12:46:13.177489] INFO: moduleinvoker: dl_convert_to_bin.v2 开始运行..

[2021-10-26 12:49:18.286543] INFO: moduleinvoker: dl_convert_to_bin.v2 运行完成[185.109061s].

[2021-10-26 12:49:18.306753] INFO: moduleinvoker: cached.v3 开始运行..

[2021-10-26 13:24:59.964824] INFO: moduleinvoker: cached.v3 运行完成[2141.658078s].

[2021-10-26 13:24:59.976056] INFO: moduleinvoker: cached.v3 开始运行..

[2021-10-26 13:25:23.372819] INFO: moduleinvoker: cached.v3 运行完成[23.396768s].

[2021-10-26 13:25:25.177196] INFO: moduleinvoker: backtest.v8 开始运行..

[2021-10-26 13:25:25.182304] INFO: backtest: biglearning backtest:V8.5.0

[2021-10-26 13:25:25.183746] INFO: backtest: product_type:stock by specified

[2021-10-26 13:25:25.306291] INFO: moduleinvoker: cached.v2 开始运行..

[2021-10-26 13:25:34.766921] INFO: backtest: 读取股票行情完成:4336032

[2021-10-26 13:25:40.770124] INFO: moduleinvoker: cached.v2 运行完成[15.463849s].

[2021-10-26 13:25:44.622145] INFO: algo: TradingAlgorithm V1.8.5

[2021-10-26 13:25:46.347660] INFO: algo: trading transform...

[2021-10-26 13:26:48.391142] INFO: Performance: Simulated 849 trading days out of 849.

[2021-10-26 13:26:48.392757] INFO: Performance: first open: 2018-01-02 09:30:00+00:00

[2021-10-26 13:26:48.394270] INFO: Performance: last close: 2021-07-01 15:00:00+00:00

[2021-10-26 13:27:03.494861] INFO: moduleinvoker: backtest.v8 运行完成[98.317669s].

[2021-10-26 13:27:03.496681] INFO: moduleinvoker: trade.v4 运行完成[100.082863s].

epoch 0   |  train_loss 1.03101|  vall_loss 0.98711|  0:03:21s
epoch 1   |  train_loss 0.98289|  vall_loss 0.98238|  0:06:42s
epoch 2   |  train_loss 0.98027|  vall_loss 0.98351|  0:10:03s
epoch 3   |  train_loss 0.97960|  vall_loss 0.98329|  0:13:24s
epoch 4   |  train_loss 0.97926|  vall_loss 0.98756|  0:16:46s
epoch 5   |  train_loss 0.97805|  vall_loss 0.98104|  0:20:07s
epoch 6   |  train_loss 0.97538|  vall_loss 0.97769|  0:23:29s
epoch 7   |  train_loss 0.97130|  vall_loss 0.97642|  0:26:49s
epoch 8   |  train_loss 0.96649|  vall_loss 0.97246|  0:30:14s
epoch 9   |  train_loss 0.96351|  vall_loss 0.97224|  0:33:37s
best loss: 0.972243 @ 9

收益率260.14%
年化收益率46.28%
基准收益率29.74%
阿尔法0.39
贝塔0.96
夏普比率1.29
胜率0.52
盈亏比1.24
收益波动率30.8%
信息比率0.09
最大回撤25.71%

bigcharts-data-start/{"__type":"tabs","__id":"bigchart-3455396ed4994e9184105b811eff7cda"}/bigcharts-data-end