克隆策略

In [23]:

# 本代码由可视化策略环境自动生成 2018年3月8日 18:31
# 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。


m3 = M.dl_layer_input.v1(
    shape='50,5',
    batch_shape='',
    dtype='float32',
    sparse=False,
    name=''
)

m11 = M.dl_layer_reshape.v1(
    inputs=m3.data,
    target_shape='50,5,1',
    name=''
)

m12 = M.dl_layer_conv2d.v1(
    inputs=m11.data,
    filters=32,
    kernel_size='3,5',
    strides='1,1',
    padding='valid',
    data_format='channels_last',
    dilation_rate='1,1',
    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=''
)

m25 = M.dl_layer_conv2d.v1(
    inputs=m12.data,
    filters=16,
    kernel_size='1,1',
    strides='1,1',
    padding='valid',
    data_format='channels_last',
    dilation_rate='1,1',
    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=''
)

m13 = M.dl_layer_reshape.v1(
    inputs=m25.data,
    target_shape='48,16,1',
    name=''
)

m26 = M.dl_layer_conv2d.v1(
    inputs=m13.data,
    filters=8,
    kernel_size='3,3',
    strides='1,1',
    padding='valid',
    data_format='channels_last',
    dilation_rate='1,1',
    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=''
)

m27 = M.dl_layer_conv2d.v1(
    inputs=m26.data,
    filters=8,
    kernel_size='3,3',
    strides='1,1',
    padding='valid',
    data_format='channels_last',
    dilation_rate='1,1',
    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=''
)

m28 = M.dl_layer_reshape.v1(
    inputs=m27.data,
    target_shape='528,8',
    name=''
)

m4 = M.dl_layer_lstm.v1(
    inputs=m28.data,
    units=8,
    activation='tanh',
    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=False,
    implementation='0',
    name=''
)

m10 = M.dl_layer_dense.v1(
    inputs=m4.data,
    units=32,
    activation='tanh',
    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=''
)

m24 = M.dl_layer_dropout.v1(
    inputs=m10.data,
    rate=0.9,
    noise_shape='',
    name=''
)

m9 = M.dl_layer_dense.v1(
    inputs=m24.data,
    units=1,
    activation='sigmoid',
    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="""(close/shift(close,1)-1)*10
(high/shift(high,1)-1)*10
(low/shift(low,1)-1)*10
(open/shift(open,1)-1)*10
(volume/shift(volume,1)-1)*10"""
)

m22 = M.instruments.v2(
    start_date='2013-01-01',
    end_date='2018-03-07',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

# 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 m21_run_bigquant_run(input_1, input_2, input_3):
    fields = ['open','high','low','close','volume']
    input_1_df = input_1.read_pickle()
    ins = input_1_df['instruments']
    start_date = input_1_df['start_date']
    end_date = input_1_df['end_date']
    df = D.history_data(ins, start_date, end_date, fields)     
    data_1 = DataSource.write_df(df)
    return Outputs(data_1=data_1, data_2=None, data_3=None)

m21 = M.cached.v3(
    input_1=m22.data,
    run=m21_run_bigquant_run
)

m1 = M.derived_feature_extractor.v2(
    input_data=m21.data_1,
    features=m8.data,
    date_col='date',
    instrument_col='instrument',
    user_functions={}
)

# 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 m14_run_bigquant_run(input_1, input_2, input_3):
    input_ds = input_1
    df = input_ds.read_df()
    df['return'] = (df.close.shift(-10)/df.close - 1)
    df['label'] = np.where(df['return'] > 0, 1, 0)
    ds = DataSource.write_df(df)
    return Outputs(data_1=ds)


m14 = M.cached.v3(
    input_1=m21.data_1,
    run=m14_run_bigquant_run
)

m15 = M.join.v3(
    data1=m14.data_1,
    data2=m1.data,
    on='date',
    how='inner',
    sort=True
)

m16 = M.dropnan.v1(
    input_data=m15.data
)

m17 = M.filter.v3(
    input_data=m16.data,
    expr='date<\'2017-10-01\'',
    output_left_data=False
)

m19 = M.dl_convert_to_bin.v1(
    input_data=m17.data,
    features=m8.data,
    window_size=50
)

m6 = M.dl_model_train.v1(
    input_model=m5.data,
    training_data=m19.data,
    optimizer='RMSprop',
    loss='binary_crossentropy',
    metrics='accuracy',
    batch_size=2048,
    epochs=30,
    n_gpus=2,
    verbose='1:输出进度条记录'
)

m18 = M.filter.v3(
    input_data=m16.data,
    expr='date>\'2017-10-01\'',
    output_left_data=False
)

m20 = M.dl_convert_to_bin.v1(
    input_data=m18.data,
    features=m8.data,
    window_size=50
)

m7 = M.dl_model_predict.v1(
    trained_model=m6.data,
    input_data=m20.data,
    batch_size=512,
    n_gpus=2,
    verbose='2:每个epoch输出一行记录'
)

# 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):
    input_series = input_1
    input_df = input_2
    test_data = input_df.read_pickle()
    pred_label = input_series.read_pickle()
  
    pred_result = pred_label.reshape(pred_label.shape[0]) 
    dt = input_3.read_df()['date'][-1*len(pred_result):]
    pred_df = pd.Series(pred_result, index=dt)
    ds = DataSource.write_df(pred_df)
    
    pred_label = np.where(pred_label>0.5,1,0)
    labels = test_data['y']
    print('准确率%s'%(np.mean(pred_label==labels)))
    
    return Outputs(data_1=ds)

m2 = M.cached.v3(
    input_1=m7.data,
    input_2=m20.data,
    input_3=m18.data,
    run=m2_run_bigquant_run
)

# 回测引擎：每日数据处理函数，每天执行一次
def m23_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.strftime('%Y-%m-%d')]
    except KeyError as e:
        return

   
    
    instrument = context.instruments[0]
    sid = context.symbol(instrument)
    cur_position = context.portfolio.positions[sid].amount
    #print('date: ',data.current_dt, '持仓： ', cur_position)
    
    # 交易逻辑
    if prediction > 0.5 and cur_position == 0:
        context.order_target_percent(context.symbol(instrument), 1)
        print(data.current_dt, '买入！')
        
    elif prediction < 0.5 and cur_position > 0:
        context.order_target_percent(context.symbol(instrument), 0)
        print(data.current_dt, '卖出！')
    
# 回测引擎：准备数据，只执行一次
def m23_prepare_bigquant_run(context):
    pass

# 回测引擎：初始化函数，只执行一次
def m23_initialize_bigquant_run(context):
    # 加载预测数据
    context.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))
# 回测引擎：每个单位时间开始前调用一次，即每日开盘前调用一次。
def m23_before_trading_start_bigquant_run(context, data):
    pass

m23 = M.trade.v3(
    instruments=m22.data,
    options_data=m2.data_1,
    start_date='2017-12-15',
    end_date='',
    handle_data=m23_handle_data_bigquant_run,
    prepare=m23_prepare_bigquant_run,
    initialize=m23_initialize_bigquant_run,
    before_trading_start=m23_before_trading_start_bigquant_run,
    volume_limit=0.025,
    order_price_field_buy='open',
    order_price_field_sell='close',
    capital_base=1000000,
    benchmark='000300.SHA',
    auto_cancel_non_tradable_orders=True,
    data_frequency='daily',
    price_type='真实价格',
    plot_charts=True,
    backtest_only=False,
    amount_integer=False
)

[2018-03-08 14:27:26.285046] INFO: bigquant: cached.v3 开始运行..
[2018-03-08 14:27:26.289621] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.290825] INFO: bigquant: cached.v3 运行完成[0.00587s].
[2018-03-08 14:27:26.296910] INFO: bigquant: input_features.v1 开始运行..
[2018-03-08 14:27:26.301760] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.303113] INFO: bigquant: input_features.v1 运行完成[0.006156s].
[2018-03-08 14:27:26.309544] INFO: bigquant: instruments.v2 开始运行..
[2018-03-08 14:27:26.312484] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.313917] INFO: bigquant: instruments.v2 运行完成[0.004358s].
[2018-03-08 14:27:26.324425] INFO: bigquant: cached.v3 开始运行..
[2018-03-08 14:27:26.327980] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.329137] INFO: bigquant: cached.v3 运行完成[0.004735s].
[2018-03-08 14:27:26.336912] INFO: bigquant: derived_feature_extractor.v2 开始运行..
[2018-03-08 14:27:26.339761] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.340932] INFO: bigquant: derived_feature_extractor.v2 运行完成[0.00402s].
[2018-03-08 14:27:26.350496] INFO: bigquant: cached.v3 开始运行..
[2018-03-08 14:27:26.353067] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.354053] INFO: bigquant: cached.v3 运行完成[0.003536s].
[2018-03-08 14:27:26.362197] INFO: bigquant: join.v3 开始运行..
[2018-03-08 14:27:26.364627] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.365781] INFO: bigquant: join.v3 运行完成[0.003574s].
[2018-03-08 14:27:26.375852] INFO: bigquant: dropnan.v1 开始运行..
[2018-03-08 14:27:26.379487] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.380869] INFO: bigquant: dropnan.v1 运行完成[0.005032s].
[2018-03-08 14:27:26.389641] INFO: bigquant: filter.v3 开始运行..
[2018-03-08 14:27:26.393387] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.394493] INFO: bigquant: filter.v3 运行完成[0.004853s].
[2018-03-08 14:27:26.407366] INFO: bigquant: dl_convert_to_bin.v1 开始运行..
[2018-03-08 14:27:26.410878] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.412038] INFO: bigquant: dl_convert_to_bin.v1 运行完成[0.004679s].
[2018-03-08 14:27:26.421284] INFO: bigquant: dl_model_train.v1 开始运行..
[2018-03-08 14:27:26.425139] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.426660] INFO: bigquant: dl_model_train.v1 运行完成[0.005376s].
[2018-03-08 14:27:26.435867] INFO: bigquant: filter.v3 开始运行..
[2018-03-08 14:27:26.439345] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.440588] INFO: bigquant: filter.v3 运行完成[0.00474s].
[2018-03-08 14:27:26.453051] INFO: bigquant: dl_convert_to_bin.v1 开始运行..
[2018-03-08 14:27:26.456667] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.458039] INFO: bigquant: dl_convert_to_bin.v1 运行完成[0.00497s].
[2018-03-08 14:27:26.466540] INFO: bigquant: dl_model_predict.v1 开始运行..
[2018-03-08 14:27:26.470485] INFO: bigquant: 命中缓存
DataSource(ad9d2da6221111e883e80242ac1100af, v2_t2)
[2018-03-08 14:27:26.472182] INFO: bigquant: dl_model_predict.v1 运行完成[0.00566s].
[2018-03-08 14:27:26.484688] INFO: bigquant: cached.v3 开始运行..
[2018-03-08 14:27:26.489003] INFO: bigquant: 命中缓存
[2018-03-08 14:27:26.490161] INFO: bigquant: cached.v3 运行完成[0.005499s].
[2018-03-08 14:27:26.524286] INFO: bigquant: backtest.v7 开始运行..
[2018-03-08 14:27:26.703875] INFO: algo: set price type:original
2018-02-12 07:00:00+00:00 买入！
[2018-03-08 14:27:29.179668] INFO: Performance: Simulated 63 trading days out of 63.
[2018-03-08 14:27:29.181095] INFO: Performance: first open: 2017-12-01 01:30:00+00:00
[2018-03-08 14:27:29.182382] INFO: Performance: last close: 2018-03-07 07:00:00+00:00

收益率11.32%
年化收益率53.54%
基准收益率0.76%
阿尔法0.49
贝塔0.23
夏普比率2.58
胜率--
盈亏比--
收益波动率19.03%
信息比率2.2
最大回撤2.36%

[2018-03-08 14:27:30.508274] INFO: bigquant: backtest.v7 运行完成[3.983995s].