复制链接

克隆策略

深度学习在期货高频上的应用¶

策略思想：¶

使用最近50分钟的价格、成交量、持仓量数据预测未来50分钟的涨跌幅。

交易标的：¶

股指期货 IF1906

模型算法：¶

LSTM深度学习算法

模型标注：¶

未来50分钟收益率

模型因子：¶

mean(open_intl,5)/amount
mean(open_intl,10)/amount
mean(open_intl,20)/amount
mean(open_intl,30)/amount
mean(open_intl,50)/amount
close/shift(close,30)
close/shift(close,20)
close/shift(close,10)
close/shift(close,5)
mean(amount,30)
max(high,10)/close 
open_intl
sum(open_intl,10)/amount 
amount/open_intl
mean(amount/open_intl,5)
mean(amount/open_intl,10)
mean(open_intl,5)

交易信号：¶

当预测值大于0.2时并且没有多仓，进场做多。
当预测值小于0.2时平掉所持有的多仓。

In [28]:

# 本代码由可视化策略环境自动生成 2022年6月24日 09:49
# 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 m20_run_bigquant_run(input_1, input_2, input_3):
     
    start_date=input_1.read_pickle()['start_date']
    end_date=input_1.read_pickle()['end_date']
    ins=input_1.read_pickle()['instruments']
    df = DataSource('bar1m_CN_FUTURE').read(instruments=ins,start_date=start_date,end_date=end_date) #bar1m_IF1906.CFE
    
    df['adjust_factor']=1.0
    data_1 = DataSource.write_df(df)
    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 m20_post_run_bigquant_run(outputs):
    return outputs

# 用户的自定义层需要写到字典中，比如
# {
#   "MyLayer": MyLayer
# }
m6_custom_objects_bigquant_run = {
    
}

# 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):
    predictions = input_1.read_pickle()
    pred_result = predictions.reshape(predictions.shape[0]) 
    dt = input_2.read_df()['date']
    pred_df = pd.Series(pred_result, index=dt)
    ds = DataSource.write_df(pred_df)
    return Outputs(data_1=ds)

# 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

# 交易引擎：初始化函数，只执行一次
def m19_initialize_bigquant_run(context):
     # 加载预测数据
    context.prediction = context.options['data'].read_df()

# 交易引擎：每个单位时间开盘前调用一次。
def m19_before_trading_start_bigquant_run(context, data):
    # 盘前处理，订阅行情等
    pass

# 交易引擎：tick数据处理函数，每个tick执行一次
def m19_handle_tick_bigquant_run(context, tick):
    pass

# 交易引擎：bar数据处理函数，每个时间单位执行一次
def m19_handle_data_bigquant_run(context, data):
    # 按日期过滤得到今日的预测数据
    try:
        prediction = context.prediction[data.current_dt]
    except KeyError as e:
        return
    instrument = context.instruments[0]
    cur_position = context.get_position(instrument, Direction.LONG)
    # 交易逻辑
    if prediction > 0.2 and cur_position.current_qty == 0:
        context.order_target(instrument, 1)
        print(data.current_dt, '买入！')
        
    elif prediction < -0.2 and cur_position.current_qty > 0:
        context.order_target(instrument, 0)
        print(data.current_dt, '卖出！')

# 交易引擎：成交回报处理函数，每个成交发生时执行一次
def m19_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 m19_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 m19_after_trading_bigquant_run(context, data):
    pass


m3 = M.dl_layer_input.v1(
    shape='50,17',
    batch_shape='',
    dtype='float32',
    sparse=False,
    name=''
)

m4 = M.dl_layer_lstm.v1(
    inputs=m3.data,
    units=32,
    activation='linear',
    recurrent_activation='hard_sigmoid',
    use_bias=True,
    kernel_initializer='glorot_uniform',
    recurrent_initializer='Orthogonal',
    bias_initializer='Ones',
    unit_forget_bias=True,
    kernel_regularizer='None',
    kernel_regularizer_l1=0,
    kernel_regularizer_l2=0,
    recurrent_regularizer='None',
    recurrent_regularizer_l1=0,
    recurrent_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',
    recurrent_constraint='None',
    bias_constraint='None',
    dropout=0,
    recurrent_dropout=0,
    return_sequences=True,
    implementation='0',
    name=''
)

m10 = M.dl_layer_dropout.v1(
    inputs=m4.data,
    rate=0.2,
    noise_shape='',
    seed=0,
    name=''
)

m25 = M.dl_layer_lstm.v1(
    inputs=m10.data,
    units=32,
    activation='sigmoid',
    recurrent_activation='hard_sigmoid',
    use_bias=True,
    kernel_initializer='glorot_uniform',
    recurrent_initializer='Orthogonal',
    bias_initializer='Zeros',
    unit_forget_bias=True,
    kernel_regularizer='None',
    kernel_regularizer_l1=0,
    kernel_regularizer_l2=0,
    recurrent_regularizer='None',
    recurrent_regularizer_l1=0,
    recurrent_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',
    recurrent_constraint='None',
    bias_constraint='None',
    dropout=0,
    recurrent_dropout=0,
    return_sequences=False,
    implementation='0',
    name=''
)

m11 = M.dl_layer_dropout.v1(
    inputs=m25.data,
    rate=0.1,
    noise_shape='',
    seed=0,
    name=''
)

m9 = M.dl_layer_dense.v1(
    inputs=m11.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=''
)

m5 = M.dl_model_init.v1(
    inputs=m3.data,
    outputs=m9.data
)

m8 = M.input_features.v1(
    features="""mean(open_intl,5)/amount
mean(open_intl,10)/amount
mean(open_intl,20)/amount
mean(open_intl,30)/amount
mean(open_intl,50)/amount
close/shift(close,30)
close/shift(close,20)
close/shift(close,10)
close/shift(close,5)
mean(amount,30)
max(high,10)/close 
open_intl
sum(open_intl,10)/amount 
amount/open_intl
mean(amount/open_intl,5)
mean(amount/open_intl,10)
mean(open_intl,5)
label= shift(close,-50)/close-1


"""
)

m24 = M.instruments.v2(
    start_date='2019-04-01',
    end_date='2019-06-22',
    market='CN_FUTURE',
    instrument_list="""IF1906.CFX
 """,
    max_count=0
)

m20 = M.cached.v3(
    input_1=m24.data,
    run=m20_run_bigquant_run,
    post_run=m20_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports=''
)

m12 = M.derived_feature_extractor.v3(
    input_data=m20.data_1,
    features=m8.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False,
    user_functions={}
)

m16 = M.filter.v3(
    input_data=m12.data,
    expr='date<\'2019-06-01\'',
    output_left_data=False
)

m28 = M.dropnan.v1(
    input_data=m16.data
)

m21 = M.filter.v3(
    input_data=m12.data,
    expr='date>=\'2019-06-01\'',
    output_left_data=False
)

m13 = M.dropnan.v1(
    input_data=m21.data
)

m15 = M.input_features.v1(
    features="""mean(open_intl,5)/amount
mean(open_intl,10)/amount
mean(open_intl,20)/amount
mean(open_intl,30)/amount
mean(open_intl,50)/amount
close/shift(close,30)
close/shift(close,20)
close/shift(close,10)
close/shift(close,5)
mean(amount,30)
max(high,10)/close 
open_intl
sum(open_intl,10)/amount 
amount/open_intl
mean(amount/open_intl,5)
mean(amount/open_intl,10)
mean(open_intl,5)"""
)

m17 = M.dl_convert_to_bin.v2(
    input_data=m13.data,
    features=m15.data,
    window_size=50,
    feature_clip=5,
    flatten=False,
    window_along_col='instrument'
)

m14 = M.dl_convert_to_bin.v2(
    input_data=m28.data,
    features=m15.data,
    window_size=50,
    feature_clip=5,
    flatten=False,
    window_along_col='instrument'
)

m6 = M.dl_model_train.v1(
    input_model=m5.data,
    training_data=m14.data,
    optimizer='RMSprop',
    loss='mean_squared_error',
    metrics='mse',
    batch_size=256,
    epochs=5,
    custom_objects=m6_custom_objects_bigquant_run,
    n_gpus=0,
    verbose='1:输出进度条记录'
)

m7 = M.dl_model_predict.v1(
    trained_model=m6.data,
    input_data=m17.data,
    batch_size=128,
    n_gpus=0,
    verbose='0:不显示'
)

m2 = M.cached.v3(
    input_1=m7.data,
    input_2=m13.data,
    run=m2_run_bigquant_run,
    post_run=m2_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports=''
)

m19 = M.hftrade.v2(
    instruments=m24.data,
    options_data=m2.data_1,
    start_date='2019-06-01',
    end_date='',
    initialize=m19_initialize_bigquant_run,
    before_trading_start=m19_before_trading_start_bigquant_run,
    handle_tick=m19_handle_tick_bigquant_run,
    handle_data=m19_handle_data_bigquant_run,
    handle_trade=m19_handle_trade_bigquant_run,
    handle_order=m19_handle_order_bigquant_run,
    after_trading=m19_after_trading_bigquant_run,
    capital_base=1000000,
    frequency='minute',
    price_type='真实价格',
    product_type='期货',
    before_start_days='0',
    order_price_field_buy='open',
    order_price_field_sell='open',
    benchmark='000300.HIX',
    plot_charts=True,
    disable_cache=False,
    replay_bdb=False,
    show_debug_info=False,
    backtest_only=False
)

[2022-06-24 09:46:20.873792] INFO: moduleinvoker: dl_layer_input.v1 运行完成[0.001986s].

[2022-06-24 09:46:21.024309] INFO: moduleinvoker: dl_layer_lstm.v1 运行完成[0.140708s].

[2022-06-24 09:46:21.040341] INFO: moduleinvoker: dl_layer_dropout.v1 运行完成[0.006091s].

[2022-06-24 09:46:21.184992] INFO: moduleinvoker: dl_layer_lstm.v1 运行完成[0.136072s].

[2022-06-24 09:46:21.199877] INFO: moduleinvoker: dl_layer_dropout.v1 运行完成[0.004809s].

[2022-06-24 09:46:21.221431] INFO: moduleinvoker: dl_layer_dense.v1 运行完成[0.010826s].

[2022-06-24 09:46:21.275633] INFO: moduleinvoker: cached.v3 开始运行..

[2022-06-24 09:46:21.290567] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.292212] INFO: moduleinvoker: cached.v3 运行完成[0.016589s].

[2022-06-24 09:46:21.298752] INFO: moduleinvoker: dl_model_init.v1 运行完成[0.066291s].

[2022-06-24 09:46:21.304276] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-06-24 09:46:21.310458] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.311984] INFO: moduleinvoker: input_features.v1 运行完成[0.007696s].

[2022-06-24 09:46:21.319319] INFO: moduleinvoker: instruments.v2 开始运行..

[2022-06-24 09:46:21.353741] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.356000] INFO: moduleinvoker: instruments.v2 运行完成[0.036667s].

[2022-06-24 09:46:21.380312] INFO: moduleinvoker: cached.v3 开始运行..

[2022-06-24 09:46:21.387844] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.390014] INFO: moduleinvoker: cached.v3 运行完成[0.009705s].

[2022-06-24 09:46:21.402905] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2022-06-24 09:46:21.410385] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.412700] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[0.009794s].

[2022-06-24 09:46:21.424647] INFO: moduleinvoker: filter.v3 开始运行..

[2022-06-24 09:46:21.467305] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.469259] INFO: moduleinvoker: filter.v3 运行完成[0.044581s].

[2022-06-24 09:46:21.486295] INFO: moduleinvoker: dropnan.v1 开始运行..

[2022-06-24 09:46:21.493472] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.495375] INFO: moduleinvoker: dropnan.v1 运行完成[0.009084s].

[2022-06-24 09:46:21.508779] INFO: moduleinvoker: filter.v3 开始运行..

[2022-06-24 09:46:21.516875] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.518807] INFO: moduleinvoker: filter.v3 运行完成[0.010029s].

[2022-06-24 09:46:21.530910] INFO: moduleinvoker: dropnan.v1 开始运行..

[2022-06-24 09:46:21.538350] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.540682] INFO: moduleinvoker: dropnan.v1 运行完成[0.009771s].

[2022-06-24 09:46:21.547664] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-06-24 09:46:21.555758] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.558426] INFO: moduleinvoker: input_features.v1 运行完成[0.01077s].

[2022-06-24 09:46:21.584609] INFO: moduleinvoker: dl_convert_to_bin.v2 开始运行..

[2022-06-24 09:46:21.594119] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.596653] INFO: moduleinvoker: dl_convert_to_bin.v2 运行完成[0.012087s].

[2022-06-24 09:46:21.618260] INFO: moduleinvoker: dl_convert_to_bin.v2 开始运行..

[2022-06-24 09:46:21.626086] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.628348] INFO: moduleinvoker: dl_convert_to_bin.v2 运行完成[0.010117s].

[2022-06-24 09:46:21.634952] INFO: moduleinvoker: dl_model_train.v1 开始运行..

[2022-06-24 09:46:21.647488] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.649432] INFO: moduleinvoker: dl_model_train.v1 运行完成[0.014484s].

[2022-06-24 09:46:21.655509] INFO: moduleinvoker: dl_model_predict.v1 开始运行..

[2022-06-24 09:46:21.664910] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.666953] INFO: moduleinvoker: dl_model_predict.v1 运行完成[0.011446s].

[2022-06-24 09:46:21.687754] INFO: moduleinvoker: cached.v3 开始运行..

[2022-06-24 09:46:21.701848] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.704236] INFO: moduleinvoker: cached.v3 运行完成[0.016495s].

[2022-06-24 09:46:21.737567] INFO: moduleinvoker: hfbacktest.v1 开始运行..

[2022-06-24 09:46:21.743662] INFO: hfbacktest: biglearning V1.4.12

[2022-06-24 09:46:21.745629] INFO: hfbacktest: bigtrader v1.9.6 2022-06-20

[2022-06-24 09:46:21.783211] INFO: moduleinvoker: cached.v2 开始运行..

[2022-06-24 09:46:21.791417] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.793948] INFO: moduleinvoker: cached.v2 运行完成[0.010751s].

[2022-06-24 09:46:21.870548] INFO: moduleinvoker: cached.v2 开始运行..

[2022-06-24 09:46:21.879211] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.881765] INFO: moduleinvoker: cached.v2 运行完成[0.011234s].

[2022-06-24 09:46:21.957791] INFO: moduleinvoker: cached.v2 开始运行..

[2022-06-24 09:46:21.964733] INFO: moduleinvoker: 命中缓存

[2022-06-24 09:46:21.971778] INFO: moduleinvoker: cached.v2 运行完成[0.013984s].

[2022-06-24 09:46:27.195148] INFO: hfbacktest: backtest done, raw_perf_ds:DataSource(a3a988b2a5b14b66b25a0fe5f3e9b8e5T)

[2022-06-24 09:46:27.755343] INFO: moduleinvoker: hfbacktest.v1 运行完成[6.017789s].

[2022-06-24 09:46:27.756962] INFO: moduleinvoker: hftrade.v2 运行完成[6.042642s].

DataSource(377e2da754ed497d95178968d56aeabfT)

2019-06-03 09:31:00 买入！
2019-06-03 09:45:00 卖出！
2019-06-03 09:50:00 买入！
2019-06-03 09:59:00 卖出！
2019-06-03 10:08:00 买入！

收益率6.04%
年化收益率187.61%
基准收益率5.56%
阿尔法0.0
贝塔1.08
夏普比率4.64
胜率0.5
盈亏比1.87
收益波动率22.67%
最大回撤2.26%

bigcharts-data-start/{"__type":"tabs","__id":"bigchart-c1d9af158b4044eea185410f7ef5167b"}/bigcharts-data-end