克隆策略

策略简介¶

因子：样例因子（7个）

因子是否标准化：是

标注：未来5日收益(不做离散化)

算法：DNN

类型：回归问题

训练集：10-15年

测试集：16-19年

选股依据：根据预测值降序排序买入

持股数：30

持仓天数：5

模型结构

输入层 7 - 因子数量

全连接层 256 激活函数为relu

dropout 0.1

全连接层 128 激活函数为relu

全连接层 1 激活函数为linear - 预测输出

In [2]:

# 本代码由可视化策略环境自动生成 2019年7月17日 17:22
# 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 m24_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[:,0], '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 m24_post_run_bigquant_run(outputs):
    return outputs

# 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 = 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 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.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 m19_prepare_bigquant_run(context):
    pass


m1 = M.instruments.v2(
    start_date='2010-01-01',
    end_date='2015-12-31',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

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/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
)

m13 = M.standardlize.v8(
    input_1=m2.data,
    columns_input='label',
    m_cached=False
)

m3 = M.input_features.v1(
    features="""close_0/mean(close_0,5)
close_0/mean(close_0,10)
close_0/mean(close_0,20)
close_0/open_0
open_0/mean(close_0,5)
open_0/mean(close_0,10)
open_0/mean(close_0,20)""",
    m_cached=False
)

m15 = M.general_feature_extractor.v7(
    instruments=m1.data,
    features=m3.data,
    start_date='',
    end_date='',
    before_start_days=0
)

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

m14 = M.standardlize.v8(
    input_1=m16.data,
    input_2=m3.data,
    columns_input='[]'
)

m7 = M.join.v3(
    data1=m13.data,
    data2=m14.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m26 = M.dl_convert_to_bin.v2(
    input_data=m7.data,
    features=m3.data,
    window_size=1,
    feature_clip=5,
    flatten=True,
    window_along_col='instrument'
)

m9 = M.instruments.v2(
    start_date=T.live_run_param('trading_date', '2016-01-01'),
    end_date=T.live_run_param('trading_date', '2019-04-20'),
    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=0
)

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

m25 = M.standardlize.v8(
    input_1=m18.data,
    input_2=m3.data,
    columns_input='[]'
)

m27 = M.dl_convert_to_bin.v2(
    input_data=m25.data,
    features=m3.data,
    window_size=1,
    feature_clip=5,
    flatten=True,
    window_along_col='instrument'
)

m6 = M.dl_layer_input.v1(
    shape='7',
    batch_shape='',
    dtype='float32',
    sparse=False,
    name=''
)

m8 = M.dl_layer_dense.v1(
    inputs=m6.data,
    units=256,
    activation='relu',
    use_bias=True,
    kernel_initializer='glorot_uniform',
    bias_initializer='Zeros',
    kernel_regularizer='None',
    kernel_regularizer_l1=0,
    kernel_regularizer_l2=0,
    bias_regularizer='None',
    bias_regularizer_l1=0,
    bias_regularizer_l2=0,
    activity_regularizer='None',
    activity_regularizer_l1=0,
    activity_regularizer_l2=0,
    kernel_constraint='None',
    bias_constraint='None',
    name=''
)

m21 = M.dl_layer_dropout.v1(
    inputs=m8.data,
    rate=0.1,
    noise_shape='',
    name=''
)

m20 = M.dl_layer_dense.v1(
    inputs=m21.data,
    units=128,
    activation='relu',
    use_bias=True,
    kernel_initializer='glorot_uniform',
    bias_initializer='Zeros',
    kernel_regularizer='None',
    kernel_regularizer_l1=0,
    kernel_regularizer_l2=0,
    bias_regularizer='None',
    bias_regularizer_l1=0,
    bias_regularizer_l2=0,
    activity_regularizer='None',
    activity_regularizer_l1=0,
    activity_regularizer_l2=0,
    kernel_constraint='None',
    bias_constraint='None',
    name=''
)

m22 = M.dl_layer_dropout.v1(
    inputs=m20.data,
    rate=0.1,
    noise_shape='',
    name=''
)

m23 = M.dl_layer_dense.v1(
    inputs=m22.data,
    units=1,
    activation='linear',
    use_bias=True,
    kernel_initializer='glorot_uniform',
    bias_initializer='Zeros',
    kernel_regularizer='None',
    kernel_regularizer_l1=0,
    kernel_regularizer_l2=0,
    bias_regularizer='None',
    bias_regularizer_l1=0,
    bias_regularizer_l2=0,
    activity_regularizer='None',
    activity_regularizer_l1=0,
    activity_regularizer_l2=0,
    kernel_constraint='None',
    bias_constraint='None',
    name=''
)

m4 = M.dl_model_init.v1(
    inputs=m6.data,
    outputs=m23.data
)

m5 = M.dl_model_train.v1(
    input_model=m4.data,
    training_data=m26.data,
    optimizer='Adam',
    loss='mean_squared_error',
    metrics='mse',
    batch_size=1024,
    epochs=5,
    n_gpus=0,
    verbose='2:每个epoch输出一行记录'
)

m11 = M.dl_model_predict.v1(
    trained_model=m5.data,
    input_data=m27.data,
    batch_size=1024,
    n_gpus=0,
    verbose='2:每个epoch输出一行记录'
)

m24 = M.cached.v3(
    input_1=m11.data,
    input_2=m18.data,
    run=m24_run_bigquant_run,
    post_run=m24_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports=''
)

m19 = M.trade.v4(
    instruments=m9.data,
    options_data=m24.data_1,
    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='后复权',
    product_type='股票',
    plot_charts=True,
    backtest_only=False,
    benchmark='000300.SHA'
)

[2019-07-17 17:05:13.341534] INFO: bigquant: instruments.v2 开始运行..

[2019-07-17 17:05:13.389137] INFO: bigquant: 命中缓存

[2019-07-17 17:05:13.390942] INFO: bigquant: instruments.v2 运行完成[0.049404s].

[2019-07-17 17:05:13.393657] INFO: bigquant: advanced_auto_labeler.v2 开始运行..

[2019-07-17 17:05:13.429060] INFO: bigquant: 命中缓存

[2019-07-17 17:05:13.431217] INFO: bigquant: advanced_auto_labeler.v2 运行完成[0.037547s].

[2019-07-17 17:05:13.433494] INFO: bigquant: standardlize.v8 开始运行..

[2019-07-17 17:05:18.530287] INFO: bigquant: standardlize.v8 运行完成[5.096649s].

[2019-07-17 17:05:18.533419] INFO: bigquant: input_features.v1 开始运行..

[2019-07-17 17:05:18.616123] INFO: bigquant: input_features.v1 运行完成[0.082686s].

[2019-07-17 17:05:18.652618] INFO: bigquant: general_feature_extractor.v7 开始运行..

[2019-07-17 17:05:18.702302] INFO: bigquant: 命中缓存

[2019-07-17 17:05:18.704840] INFO: bigquant: general_feature_extractor.v7 运行完成[0.052213s].

[2019-07-17 17:05:18.708448] INFO: bigquant: derived_feature_extractor.v3 开始运行..

[2019-07-17 17:05:23.165191] INFO: derived_feature_extractor: 提取完成 close_0/mean(close_0,5), 3.209s

[2019-07-17 17:05:26.124895] INFO: derived_feature_extractor: 提取完成 close_0/mean(close_0,10), 2.957s

[2019-07-17 17:05:29.096943] INFO: derived_feature_extractor: 提取完成 close_0/mean(close_0,20), 2.966s

[2019-07-17 17:05:29.112378] INFO: derived_feature_extractor: 提取完成 close_0/open_0, 0.013s

[2019-07-17 17:05:32.715614] INFO: derived_feature_extractor: 提取完成 open_0/mean(close_0,5), 3.601s

[2019-07-17 17:05:35.424654] INFO: derived_feature_extractor: 提取完成 open_0/mean(close_0,10), 2.707s

[2019-07-17 17:05:38.332868] INFO: derived_feature_extractor: 提取完成 open_0/mean(close_0,20), 2.906s

[2019-07-17 17:05:39.056880] INFO: derived_feature_extractor: /y_2010, 431567

[2019-07-17 17:05:39.627020] INFO: derived_feature_extractor: /y_2011, 511455

[2019-07-17 17:05:40.291014] INFO: derived_feature_extractor: /y_2012, 565675

[2019-07-17 17:05:40.986461] INFO: derived_feature_extractor: /y_2013, 564168

[2019-07-17 17:05:41.679564] INFO: derived_feature_extractor: /y_2014, 569948

[2019-07-17 17:05:42.370702] INFO: derived_feature_extractor: /y_2015, 569698

[2019-07-17 17:05:43.638047] INFO: bigquant: derived_feature_extractor.v3 运行完成[24.929589s].

[2019-07-17 17:05:43.642173] INFO: bigquant: standardlize.v8 开始运行..

[2019-07-17 17:05:58.351563] INFO: bigquant: standardlize.v8 运行完成[14.70937s].

[2019-07-17 17:05:58.355068] INFO: bigquant: join.v3 开始运行..

[2019-07-17 17:06:08.243425] INFO: join: /data, 行数=3133725/3159055, 耗时=8.936639s

[2019-07-17 17:06:10.813906] INFO: join: 最终行数: 3133725

[2019-07-17 17:06:10.817953] INFO: bigquant: join.v3 运行完成[12.462869s].

[2019-07-17 17:06:10.858171] INFO: bigquant: dl_convert_to_bin.v2 开始运行..

[2019-07-17 17:06:18.035672] INFO: bigquant: dl_convert_to_bin.v2 运行完成[7.17748s].

[2019-07-17 17:06:18.039394] INFO: bigquant: instruments.v2 开始运行..

[2019-07-17 17:06:18.206988] INFO: bigquant: 命中缓存

[2019-07-17 17:06:18.209039] INFO: bigquant: instruments.v2 运行完成[0.169641s].

[2019-07-17 17:06:18.266745] INFO: bigquant: general_feature_extractor.v7 开始运行..

[2019-07-17 17:06:18.324282] INFO: bigquant: 命中缓存

[2019-07-17 17:06:18.326403] INFO: bigquant: general_feature_extractor.v7 运行完成[0.059664s].

[2019-07-17 17:06:18.329452] INFO: bigquant: derived_feature_extractor.v3 开始运行..

[2019-07-17 17:06:22.861225] INFO: derived_feature_extractor: 提取完成 close_0/mean(close_0,5), 3.426s

[2019-07-17 17:06:26.808928] INFO: derived_feature_extractor: 提取完成 close_0/mean(close_0,10), 3.945s

[2019-07-17 17:06:30.482736] INFO: derived_feature_extractor: 提取完成 close_0/mean(close_0,20), 3.671s

[2019-07-17 17:06:30.494051] INFO: derived_feature_extractor: 提取完成 close_0/open_0, 0.009s

[2019-07-17 17:06:34.459094] INFO: derived_feature_extractor: 提取完成 open_0/mean(close_0,5), 3.963s

[2019-07-17 17:06:38.488371] INFO: derived_feature_extractor: 提取完成 open_0/mean(close_0,10), 4.027s

[2019-07-17 17:06:42.295940] INFO: derived_feature_extractor: 提取完成 open_0/mean(close_0,20), 3.805s

[2019-07-17 17:06:42.972644] INFO: derived_feature_extractor: /y_2016, 641546

[2019-07-17 17:06:43.714168] INFO: derived_feature_extractor: /y_2017, 743233

[2019-07-17 17:06:44.948682] INFO: derived_feature_extractor: /y_2018, 816987

[2019-07-17 17:06:46.001522] INFO: derived_feature_extractor: /y_2019, 256890

[2019-07-17 17:06:47.533112] INFO: bigquant: derived_feature_extractor.v3 运行完成[29.203629s].

[2019-07-17 17:06:47.538391] INFO: bigquant: standardlize.v8 开始运行..

[2019-07-17 17:06:57.856758] INFO: bigquant: standardlize.v8 运行完成[10.318337s].

[2019-07-17 17:06:57.997183] INFO: bigquant: dl_convert_to_bin.v2 开始运行..

[2019-07-17 17:07:00.527888] INFO: bigquant: dl_convert_to_bin.v2 运行完成[2.530687s].

[2019-07-17 17:07:00.875431] INFO: bigquant: cached.v3 开始运行..

[2019-07-17 17:07:00.989644] INFO: bigquant: 命中缓存

[2019-07-17 17:07:00.992010] INFO: bigquant: cached.v3 运行完成[0.116581s].

[2019-07-17 17:07:01.714357] INFO: dl_model_train: 准备训练，训练样本个数：3133725，迭代次数：5

[2019-07-17 17:14:15.148830] INFO: dl_model_train: 训练结束，耗时：433.43s

[2019-07-17 17:14:15.487248] INFO: bigquant: dl_model_train.v1 运行完成[434.492177s].

[2019-07-17 17:14:28.214569] INFO: bigquant: dl_model_predict.v1 运行完成[12.720695s].

[2019-07-17 17:14:28.219610] INFO: bigquant: cached.v3 开始运行..

[2019-07-17 17:14:34.218014] INFO: bigquant: cached.v3 运行完成[5.998392s].

[2019-07-17 17:14:34.283409] INFO: bigquant: backtest.v8 开始运行..

[2019-07-17 17:14:34.287449] INFO: bigquant: biglearning backtest:V8.2.5

[2019-07-17 17:14:34.290270] INFO: bigquant: product_type:stock by specified

[2019-07-17 17:14:34.570313] INFO: bigquant: cached.v2 开始运行..

[2019-07-17 17:14:34.611992] INFO: bigquant: 命中缓存

[2019-07-17 17:14:34.614649] INFO: bigquant: cached.v2 运行完成[0.044335s].

[2019-07-17 17:14:54.367153] INFO: algo: TradingAlgorithm V1.5.1

[2019-07-17 17:14:57.323512] INFO: algo: trading transform...

[2019-07-17 17:16:13.420810] INFO: Performance: Simulated 803 trading days out of 803.

[2019-07-17 17:16:13.422856] INFO: Performance: first open: 2016-01-04 09:30:00+00:00

[2019-07-17 17:16:13.424397] INFO: Performance: last close: 2019-04-19 15:00:00+00:00

[2019-07-17 17:16:34.652101] INFO: bigquant: backtest.v8 运行完成[120.368751s].

Epoch 1/5
 - 84s - loss: 0.9901 - mean_squared_error: 0.9901
Epoch 2/5
 - 85s - loss: 0.9891 - mean_squared_error: 0.9891
Epoch 3/5
 - 95s - loss: 0.9889 - mean_squared_error: 0.9889
Epoch 4/5
 - 83s - loss: 0.9886 - mean_squared_error: 0.9886
Epoch 5/5
 - 86s - loss: 0.9885 - mean_squared_error: 0.9885

DataSource(80bec2c1c82e484da700d5e27e95fa50T, v3)

收益率122.94%
年化收益率28.61%
基准收益率10.44%
阿尔法0.25
贝塔0.74
夏普比率0.94
胜率0.58
盈亏比0.94
收益波动率27.58%
信息比率0.06
最大回撤20.37%

bigcharts-data-start/{"__id":"bigchart-cf7a8d91609c48c6b04a86a717f35e80","__type":"tabs"}/bigcharts-data-end