克隆策略

In [ ]:

In [4]:

# 本代码由可视化策略环境自动生成 2021年9月5日 14:28
# 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 m12_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.preprocessing import MinMaxScaler
    df = input_1.read()
    
    
    
    
    data_1 = DataSource.write_df(df)
    data_2 = DataSource.write_pickle(df)
    return Outputs(data_1=data_1, data_2=data_2, 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 m12_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):

    context.ranker_prediction = context.options['data'].read_df()
    context.set_commission(PerOrder(buy_cost=0.0003, sell_cost=0.0013, min_cost=5))
    context.stock_count = 5
    context.hold_days = 15

# 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):
    
    if context.trading_day_index % context.hold_days != 0:
        return 
    
    # 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')]
    
    stock_to_buy = list(ranker_prediction.instrument)[:context.stock_count]
    # 通过positions对象，使用列表生成式的方法获取目前持仓的股票列表
    stock_hold_now = [equity.symbol for equity in context.portfolio.positions]
    # 继续持有的股票：调仓时，如果买入的股票已经存在于目前的持仓里，那么应继续持有
    no_need_to_sell = [i for i in stock_hold_now if i in stock_to_buy]
    # 需要卖出的股票
    stock_to_sell = [i for i in stock_hold_now if i not in no_need_to_sell]
    
    assert type(stock_to_buy) == list, '格式类型不对！'
    
    
    # 卖出
    for stock in stock_to_sell:
        if data.can_trade(context.symbol(stock)):
            # order_target_percent是平台的一个下单接口，表明下单使得该股票的权重为0，
            #   即卖出全部股票，可参考回测文档
            context.order_target_percent(context.symbol(stock), 0)
    
    # 如果当天没有买入的股票，就返回
    if len(stock_to_buy) == 0:
        return

     
    context.stock_weights = T.norm([1 / math.log(i + 2) for i in range(0, len(stock_to_buy))])

    # 买入
    c = 0
    for stock in stock_to_buy:
        
        if data.can_trade(context.symbol(stock)):
            # 下单使得某只股票的持仓权重达到weight，因为
            # weight大于0,因此是等权重买入
            context.order_target_percent(context.symbol(stock), context.stock_weights[c])
        c += 1
# 回测引擎：准备数据，只执行一次
def m19_prepare_bigquant_run(context):
    pass


g = T.Graph({

    'm1': 'M.instruments.v2',
    'm1.start_date': '2018-01-01',
    'm1.end_date': '2021-01-01',
    'm1.market': 'CN_STOCK_A',
    'm1.instrument_list': '',
    'm1.max_count': 0,

    'm2': 'M.advanced_auto_labeler.v2',
    'm2.instruments': T.Graph.OutputPort('m1.data'),
    'm2.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, -15) / shift(open, -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))

# 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)
""",
    'm2.start_date': '',
    'm2.end_date': '',
    'm2.benchmark': '000300.SHA',
    'm2.drop_na_label': True,
    'm2.cast_label_int': True,

    'm3': 'M.input_features.v1',
    'm3.features': """open_0/mean(close_0,10)
amount_0/mean(amount_0,3)
ta_bbands_lowerband_14_0
close_0/mean(close_0,3)
open_0/mean(close_0,3)
(close_0-open_0)/close_1
ts_argmin(low_0,20)
rank(std(amount_0,15))
return_1/return_5
turn_0/mean(turn_0,3)""",

    'm15': 'M.general_feature_extractor.v7',
    'm15.instruments': T.Graph.OutputPort('m1.data'),
    'm15.features': T.Graph.OutputPort('m3.data'),
    'm15.start_date': '',
    'm15.end_date': '',
    'm15.before_start_days': 0,

    'm16': 'M.derived_feature_extractor.v3',
    'm16.input_data': T.Graph.OutputPort('m15.data'),
    'm16.features': T.Graph.OutputPort('m3.data'),
    'm16.date_col': 'date',
    'm16.instrument_col': 'instrument',
    'm16.drop_na': False,
    'm16.remove_extra_columns': False,

    'm7': 'M.join.v3',
    'm7.data1': T.Graph.OutputPort('m2.data'),
    'm7.data2': T.Graph.OutputPort('m16.data'),
    'm7.on': 'date,instrument',
    'm7.how': 'inner',
    'm7.sort': False,

    'm6': 'M.dropnan.v2',
    'm6.input_data': T.Graph.OutputPort('m7.data'),

    'm12': 'M.cached.v3',
    'm12.input_1': T.Graph.OutputPort('m16.data'),
    'm12.run': m12_run_bigquant_run,
    'm12.post_run': m12_post_run_bigquant_run,
    'm12.input_ports': '',
    'm12.params': '{}',
    'm12.output_ports': '',

    'm5': 'M.stock_ranker_train.v6',
    'm5.training_ds': T.Graph.OutputPort('m6.data'),
    'm5.features': T.Graph.OutputPort('m3.data'),
    'm5.learning_algorithm': '排序',
    'm5.number_of_leaves': 30,
    'm5.minimum_docs_per_leaf': 1000,
    'm5.number_of_trees': 20,
    'm5.learning_rate': 0.1,
    'm5.max_bins': 1023,
    'm5.feature_fraction': 1,
    'm5.data_row_fraction': 1,
    'm5.ndcg_discount_base': 1,
    'm5.m_lazy_run': False,

    'm9': 'M.instruments.v2',
    'm9.start_date': T.live_run_param('trading_date', '2021-01-01'),
    'm9.end_date': T.live_run_param('trading_date', '2021-06-01'),
    'm9.market': 'CN_STOCK_A',
    'm9.instrument_list': '',
    'm9.max_count': 0,

    'm17': 'M.general_feature_extractor.v7',
    'm17.instruments': T.Graph.OutputPort('m9.data'),
    'm17.features': T.Graph.OutputPort('m3.data'),
    'm17.start_date': '',
    'm17.end_date': '',
    'm17.before_start_days': 0,

    'm18': 'M.derived_feature_extractor.v3',
    'm18.input_data': T.Graph.OutputPort('m17.data'),
    'm18.features': T.Graph.OutputPort('m3.data'),
    'm18.date_col': 'date',
    'm18.instrument_col': 'instrument',
    'm18.drop_na': False,
    'm18.remove_extra_columns': False,

    'm10': 'M.dropnan.v2',
    'm10.input_data': T.Graph.OutputPort('m18.data'),

    'm8': 'M.stock_ranker_predict.v5',
    'm8.model': T.Graph.OutputPort('m5.model'),
    'm8.data': T.Graph.OutputPort('m10.data'),
    'm8.m_lazy_run': False,

    'm19': 'M.trade.v4',
    'm19.instruments': T.Graph.OutputPort('m9.data'),
    'm19.options_data': T.Graph.OutputPort('m8.predictions'),
    'm19.start_date': '',
    'm19.end_date': '',
    'm19.initialize': m19_initialize_bigquant_run,
    'm19.handle_data': m19_handle_data_bigquant_run,
    'm19.prepare': m19_prepare_bigquant_run,
    'm19.volume_limit': 0.025,
    'm19.order_price_field_buy': 'open',
    'm19.order_price_field_sell': 'open',
    'm19.capital_base': 3000000,
    'm19.auto_cancel_non_tradable_orders': True,
    'm19.data_frequency': 'daily',
    'm19.price_type': '后复权',
    'm19.product_type': '股票',
    'm19.plot_charts': True,
    'm19.backtest_only': False,
    'm19.benchmark': '',
})

# g.run({})


def m4_run_bigquant_run(bq_graph, inputs):
    factor_pool =['rank(std(amount_0,15))','rank_avg_amount_0/rank_avg_amount_8',
                    'ts_argmin(low_0,20)','rank_return_30','(low_1-close_0)/close_0','ta_bbands_lowerband_14_0',
                    'mean(mf_net_pct_s_0,4)','amount_0/avg_amount_3','return_0/return_5','return_1/return_5',
                    'rank_avg_amount_7/rank_avg_amount_10','ta_sma_10_0/close_0','sqrt(high_0*low_0)-amount_0/volume_0*adjust_factor_0',
                    'avg_turn_15/turn_0','return_10','mf_net_pct_s_0','(close_0-open_0)/close_1',
                    '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','close_0/open_0','close_0/mean(close_0,3)',
                    'close_0/mean(close_0,5)','close_0/mean(close_0,15)','close_0/mean(close_0,30)','close_0/mean(close_0,60)',
                    'amount_0/mean(amount_0,3)','amount_0/mean(amount_0,5)','amount_0/mean(amount_0,10)','amount_0/mean(amount_0,15)',
                    'amount_0/mean(amount_0,30)','amount_0/mean(amount_0,60)','turn_0/mean(turn_0,3)','turn_0/mean(turn_0,5)',
                    'turn_0/mean(turn_0,10)','turn_0/mean(turn_0,15)','turn_0/mean(turn_0,30)','turn_0/mean(turn_0,60)','open_0/mean(close_0,3)',
                    'open_0/mean(close_0,5)','open_0/mean(close_0,10)','open_0/mean(close_0,15)','open_0/mean(close_0,30)','open_0/mean(close_0,60)']

    batch_num = 10 # 多少组,需要跑多少组策略
    batch_factor = list()
    for i in range(batch_num):
        random.seed(i)
        factor_num = 15 # 每组多少个因子
        batch_factor.append(random.sample(factor_pool, factor_num))

    parameters_list = []
     
    for feature in  batch_factor:
        parameters = {'m3.features': '\n'.join(feature)}
        parameters_list.append({'parameters': parameters})
        
    
    def run(parameters):
        try:
            return g.run(parameters)
        except Exception as e:
            print('ERROR --------', e)
            return None
 
    results = T.parallel_map(run, parameters_list, max_workers=10, remote_run=True, silent=True)  # 任务数  # 是否远程  # 

    return results


m4 = M.hyper_run.v1(
    run=m4_run_bigquant_run,
    run_now=True,
    bq_graph=g
)

[2021-09-05 14:25:29.700571] WARNING: AI: 您最多可同时运行2个分布式AI任务，升级会员/开通资源扩展包以获取更多的AI任务位 [url="https://bigquant.com/account/big_member/?from=lab1" style="display: inline-block;padding: 5px 7px;border-radius: 2px;background: #F0BC41;color: white"]升级会员/开通资源拓展包[/url]

[2021-09-05 14:25:29.702085] INFO: AI: 开始并行运算, remote_run=True, workers=2 ..

[2021-09-05 14:25:29.703299] INFO: AI: [ParallelEx(n_jobs=2)]: Using backend ThreadingBackend with 2 concurrent workers.

[2021-09-05 14:25:29.832843] INFO: cached.v2.10ecfeaa: 任务状态: Pending

[2021-09-05 14:25:29.841520] INFO: cached.v2.10ece2ee: 任务状态: Pending

[2021-09-05 14:25:39.865102] INFO: cached.v2.10ecfeaa: 任务状态: Running

[2021-09-05 14:25:39.872074] INFO: cached.v2.10ece2ee: 任务状态: Running

[2021-09-05 14:25:49.907081] INFO: cached.v2.10ece2ee: 任务状态: Succeeded

[2021-09-05 14:25:49.911769] INFO: AI: [ParallelEx(n_jobs=2)]: Done   1 tasks      | elapsed:   20.2s

[2021-09-05 14:25:49.999368] INFO: cached.v2.1cf131e4: 任务状态: Pending

[2021-09-05 14:25:59.924999] INFO: cached.v2.10ecfeaa: 任务状态: Succeeded

[2021-09-05 14:25:59.928597] INFO: AI: [ParallelEx(n_jobs=2)]: Done   2 tasks      | elapsed:   30.2s

[2021-09-05 14:26:00.004987] INFO: cached.v2.22e9a798: 任务状态: Pending

[2021-09-05 14:26:00.037396] INFO: cached.v2.1cf131e4: 任务状态: Succeeded

[2021-09-05 14:26:00.040607] INFO: AI: [ParallelEx(n_jobs=2)]: Done   3 tasks      | elapsed:   30.3s

[2021-09-05 14:26:00.112834] INFO: cached.v2.22fadeb4: 任务状态: Pending

[2021-09-05 14:26:10.034695] INFO: cached.v2.22e9a798: 任务状态: Running

[2021-09-05 14:26:10.142217] INFO: cached.v2.22fadeb4: 任务状态: Succeeded

[2021-09-05 14:26:10.145926] INFO: AI: [ParallelEx(n_jobs=2)]: Done   4 tasks      | elapsed:   40.4s

[2021-09-05 14:26:10.227542] INFO: cached.v2.2901d01a: 任务状态: Pending

[2021-09-05 14:26:20.065769] INFO: cached.v2.22e9a798: 任务状态: Succeeded

[2021-09-05 14:26:20.069514] INFO: AI: [ParallelEx(n_jobs=2)]: Done   5 tasks      | elapsed:   50.4s

[2021-09-05 14:26:20.140515] INFO: cached.v2.2eeb198c: 任务状态: Pending

[2021-09-05 14:26:20.254446] INFO: cached.v2.2901d01a: 任务状态: Running

[2021-09-05 14:26:30.178566] INFO: cached.v2.2eeb198c: 任务状态: Running

[2021-09-05 14:26:30.284778] INFO: cached.v2.2901d01a: 任务状态: Succeeded

[2021-09-05 14:26:30.288111] INFO: AI: [ParallelEx(n_jobs=2)]: Done   6 tasks      | elapsed:  1.0min

[2021-09-05 14:26:30.358519] INFO: cached.v2.3502317a: 任务状态: Pending

[2021-09-05 14:26:40.385979] INFO: cached.v2.3502317a: 任务状态: Succeeded

[2021-09-05 14:26:40.389468] INFO: AI: [ParallelEx(n_jobs=2)]: Done   7 tasks      | elapsed:  1.2min

[2021-09-05 14:26:40.499110] INFO: cached.v2.3b07c4fe: 任务状态: Pending

[2021-09-05 14:26:50.234357] INFO: cached.v2.2eeb198c: 任务状态: Succeeded

[2021-09-05 14:26:50.237506] INFO: AI: [ParallelEx(n_jobs=2)]: Done   8 out of  10 | elapsed:  1.3min remaining:   20.1s

[2021-09-05 14:26:50.339050] INFO: cached.v2.40e8d2e6: 任务状态: Pending

[2021-09-05 14:26:50.529625] INFO: cached.v2.3b07c4fe: 任务状态: Succeeded

[2021-09-05 14:27:00.369848] INFO: cached.v2.40e8d2e6: 任务状态: Running

[2021-09-05 14:27:10.397440] INFO: cached.v2.40e8d2e6: 任务状态: Succeeded

[2021-09-05 14:27:10.400602] INFO: AI: [ParallelEx(n_jobs=2)]: Done  10 out of  10 | elapsed:  1.7min remaining:    0.0s

[2021-09-05 14:27:10.402194] INFO: AI: [ParallelEx(n_jobs=2)]: Done  10 out of  10 | elapsed:  1.7min finished

[2021-09-05 14:27:10.404175] INFO: moduleinvoker: hyper_run.v1 运行完成[100.883964s].

In [7]:

m4.result

Out[7]:

[None, None, None, None, None, None, None, None, None, None]

In [ ]:

len(m4.result)

In [ ]:

# 查看第一个任务返回结果中的预测模块m8的预测结果前5条
m4.result[0]['m8'].predictions.read_df().head()  # 每一个任务  每一个模块 的输入

In [ ]:

# 绘制第一个任务返回结果中回测模块的绩效曲线
m4.result[0]['m19'].display()