复制链接

克隆策略

In [5]:

# 本代码由可视化策略环境自动生成 2023年3月27日 22:30
# 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。


# 回测引擎：初始化函数，只执行一次
def m4_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 = 5
    # 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 = 0.5
    context.options['hold_days'] = 2

# 回测引擎：每日数据处理函数，每天执行一次
def m4_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')]
    buy_condition = 2.0
    sell_condition = 0.8

    # 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, (0.5 if is_staging else 0.75) * 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StockRanker预测的排序末位淘汰
    if True:
        equities = {e.symbol: e for e, p in context.perf_tracker.position_tracker.positions.items()}
        instruments12 = list(ranker_prediction.instrument[ranker_prediction.instrument.apply(
                lambda x: x in equities )])
        position_prediction = ranker_prediction[ranker_prediction.instrument.isin(instruments12)]
        instruments = list(position_prediction.instrument[position_prediction.score < sell_condition])
        for instrument in instruments:
            context.order_target(context.symbol(instrument), 0)
            

    # 3. 生成买入订单：按StockRanker预测的排序，买入前面的stock_count只股票
    rank_buy = ranker_prediction[ranker_prediction.score>buy_condition]
    buy_instruments1 = list(rank_buy.instrument)
    buy_scores1 = list(rank_buy.score)
    buy_instruments = buy_instruments1[:np.where(len(buy_instruments1)>5,5,len(buy_instruments1))]
    buy_scores = buy_scores1[:len(buy_instruments)]
    buy_cash_weights = buy_scores/np.sum(buy_scores)
    max_cash_per_instrument = context.portfolio.portfolio_value * context.max_cash_per_instrument
    if any(buy_instruments):
        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)
            today_price = data.current(symbol(instrument), ['amount', 'volume'])
            buy_price =today_price['amount'] / today_price['volume']
            buy_amount = int(round(cash/(buy_price*100)))
            if buy_amount > 0: 
                context.order_lots(symbol(instrument),buy_amount)
# 回测引擎：准备数据，只执行一次
def m4_prepare_bigquant_run(context):
    pass

m1 = M.instruments.v2(
    start_date='2020-01-01',
    end_date='2022-04-01',
    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)
#signedpower((shift(close, -5) / shift(open, -1)-1),log10(market_cap_float)/pe_ttm)
#where((shift(close, -3) / shift(open, -1) > 0) & (correlation(close, amount, -3)>0),correlation(close, amount, -3) ,0)
(shift(close, -10) / shift(open, -1)-1)*100
# 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, 60)

# 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=True
)

m3 = M.input_features.v1(
    features="""Alpha_1=-1*rank((ts_max(high_0,20)+ts_min(low_0,20)-open_20-close_0))
Alpha_2=-1*rank((close_0-open_20)/(ts_max(high_0,20)-ts_min(low_0,20)))
Alpha_3=rank(-1*(ts_max(close_0,20)-close_0)/(ts_max(close_0,20)-ts_min(close_0,20)))

Alpha_4=rank(sum(amount_0*sign(max(open_0+close_0-high_0-low_0,close_0-open_0)/max(high_0-close_1,high_0-low_0,close_1-low_0)-0.1),20))
Alpha_5=rank(return_20)

"""
)

m15 = M.general_feature_extractor.v7(
    instruments=m1.data,
    features=m3.data,
    start_date='',
    end_date='',
    before_start_days=120
)

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

m7 = M.join.v3(
    data1=m2.data,
    data2=m16.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m11 = M.chinaa_stock_filter.v1(
    input_data=m7.data,
    index_constituent_cond=['全部'],
    board_cond=['上证主板', '深证主板', '创业板'],
    industry_cond=['全部'],
    st_cond=['正常'],
    delist_cond=['非退市'],
    output_left_data=False
)

m6 = M.dropnan.v2(
    input_data=m11.data
)

m12 = M.filter.v3(
    input_data=m6.data,
    expr='date<\'2017-01-01\'',
    output_left_data=True
)

m5 = M.features_short.v1(
    input_1=m3.data
)

m20 = M.stock_ranker_train.v6(
    training_ds=m12.data,
    features=m5.data_1,
    test_ds=m12.left_data,
    learning_algorithm='排序',
    number_of_leaves=70,
    minimum_docs_per_leaf=2400,
    number_of_trees=60,
    learning_rate=0.15,
    max_bins=800,
    feature_fraction=1,
    data_row_fraction=1,
    plot_charts=True,
    ndcg_discount_base=1,
    m_lazy_run=False
)

m9 = M.instruments.v2(
    start_date=T.live_run_param('trading_date', '2022-04-02'),
    end_date=T.live_run_param('trading_date', '2023-04-01'),
    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=120
)

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

m13 = M.chinaa_stock_filter.v1(
    input_data=m18.data,
    index_constituent_cond=['全部'],
    board_cond=['上证主板', '深证主板', '创业板'],
    industry_cond=['全部'],
    st_cond=['正常'],
    delist_cond=['非退市'],
    output_left_data=False
)

m10 = M.dropnan.v2(
    input_data=m13.data
)

m8 = M.stock_ranker_predict.v5(
    model=m20.model,
    data=m10.data,
    m_lazy_run=False
)

m4 = M.trade.v4(
    instruments=m9.data,
    options_data=m8.predictions,
    start_date='',
    end_date='',
    initialize=m4_initialize_bigquant_run,
    handle_data=m4_handle_data_bigquant_run,
    prepare=m4_prepare_bigquant_run,
    volume_limit=0,
    order_price_field_buy='open',
    order_price_field_sell='close',
    capital_base=100000,
    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=''
)

[2023-03-27 22:25:05.012613] INFO: moduleinvoker: instruments.v2 开始运行..

[2023-03-27 22:25:05.024357] INFO: moduleinvoker: 命中缓存

[2023-03-27 22:25:05.026979] INFO: moduleinvoker: instruments.v2 运行完成[0.014372s].

[2023-03-27 22:25:05.039002] INFO: moduleinvoker: advanced_auto_labeler.v2 开始运行..

[2023-03-27 22:25:09.505090] INFO: 自动标注(股票): 加载历史数据: 2285675 行

[2023-03-27 22:25:09.506888] INFO: 自动标注(股票): 开始标注 ..

[2023-03-27 22:25:12.538208] INFO: moduleinvoker: advanced_auto_labeler.v2 运行完成[7.499202s].

[2023-03-27 22:25:12.544503] INFO: moduleinvoker: input_features.v1 开始运行..

[2023-03-27 22:25:12.552147] INFO: moduleinvoker: 命中缓存

[2023-03-27 22:25:12.553632] INFO: moduleinvoker: input_features.v1 运行完成[0.009154s].

[2023-03-27 22:25:12.572155] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2023-03-27 22:25:13.606275] INFO: 基础特征抽取: 年份 2019, 特征行数=295265

[2023-03-27 22:25:16.336881] INFO: 基础特征抽取: 年份 2020, 特征行数=945961

[2023-03-27 22:25:19.148651] INFO: 基础特征抽取: 年份 2021, 特征行数=1061527

[2023-03-27 22:25:20.190304] INFO: 基础特征抽取: 年份 2022, 特征行数=278188

[2023-03-27 22:25:20.366233] INFO: 基础特征抽取: 总行数: 2580941

[2023-03-27 22:25:20.372317] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[7.800169s].

[2023-03-27 22:25:20.380411] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2023-03-27 22:25:31.682742] INFO: derived_feature_extractor: 提取完成 Alpha_1=-1*rank((ts_max(high_0,20)+ts_min(low_0,20)-open_20-close_0)), 5.968s

[2023-03-27 22:25:37.521656] INFO: derived_feature_extractor: 提取完成 Alpha_2=-1*rank((close_0-open_20)/(ts_max(high_0,20)-ts_min(low_0,20))), 5.837s

[2023-03-27 22:25:45.813397] INFO: derived_feature_extractor: 提取完成 Alpha_3=rank(-1*(ts_max(close_0,20)-close_0)/(ts_max(close_0,20)-ts_min(close_0,20))), 8.290s

[2023-03-27 22:25:49.752452] INFO: derived_feature_extractor: 提取完成 Alpha_4=rank(sum(amount_0*sign(max(open_0+close_0-high_0-low_0,close_0-open_0)/max(high_0-close_1,high_0-low_0,close_1-low_0)-0.1),20)), 3.936s

[2023-03-27 22:25:51.140958] INFO: derived_feature_extractor: 提取完成 Alpha_5=rank(return_20), 1.387s

[2023-03-27 22:25:52.079072] INFO: derived_feature_extractor: /y_2019, 295265

[2023-03-27 22:25:53.773046] INFO: derived_feature_extractor: /y_2020, 945961

[2023-03-27 22:25:55.910152] INFO: derived_feature_extractor: /y_2021, 1061527

[2023-03-27 22:25:56.965188] INFO: derived_feature_extractor: /y_2022, 278188

[2023-03-27 22:25:57.377972] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[36.997543s].

[2023-03-27 22:25:57.387678] INFO: moduleinvoker: join.v3 开始运行..

[2023-03-27 22:26:02.682839] INFO: join: /y_2019, 行数=0/295265, 耗时=1.118925s

[2023-03-27 22:26:05.843801] INFO: join: /y_2020, 行数=939839/945961, 耗时=3.158141s

[2023-03-27 22:26:09.464832] INFO: join: /y_2021, 行数=1058488/1061527, 耗时=3.613384s

[2023-03-27 22:26:10.830512] INFO: join: /y_2022, 行数=230002/278188, 耗时=1.35723s

[2023-03-27 22:26:10.974612] INFO: join: 最终行数: 2228329

[2023-03-27 22:26:10.993265] INFO: moduleinvoker: join.v3 运行完成[13.605583s].

[2023-03-27 22:26:11.003530] INFO: moduleinvoker: chinaa_stock_filter.v1 开始运行..

[2023-03-27 22:26:17.328670] INFO: A股股票过滤: 过滤 /y_2020, 865646/0/939839

[2023-03-27 22:26:23.548970] INFO: A股股票过滤: 过滤 /y_2021, 938815/0/1058488

[2023-03-27 22:26:25.026253] INFO: A股股票过滤: 过滤 /y_2022, 198532/0/230002

[2023-03-27 22:26:25.032894] INFO: A股股票过滤: 过滤完成, 2002993 + 0

[2023-03-27 22:26:25.055510] INFO: moduleinvoker: chinaa_stock_filter.v1 运行完成[14.051973s].

[2023-03-27 22:26:25.064892] INFO: moduleinvoker: dropnan.v2 开始运行..

[2023-03-27 22:26:28.142019] INFO: dropnan: /y_2020, 861606/865646

[2023-03-27 22:26:30.590864] INFO: dropnan: /y_2021, 932514/938815

[2023-03-27 22:26:31.130867] INFO: dropnan: /y_2022, 197620/198532

[2023-03-27 22:26:31.264112] INFO: dropnan: 行数: 1991740/2002993

[2023-03-27 22:26:31.270475] INFO: moduleinvoker: dropnan.v2 运行完成[6.205572s].

[2023-03-27 22:26:31.280582] INFO: moduleinvoker: filter.v3 开始运行..

[2023-03-27 22:26:31.303107] INFO: filter: 使用表达式 date[2023-03-27 22:26:33.867032] INFO: filter: 过滤 /y_2020, 0/861606/861606
[2023-03-27 22:26:36.119876] INFO: filter: 过滤 /y_2021, 0/932514/932514
[2023-03-27 22:26:36.619118] INFO: filter: 过滤 /y_2022, 0/197620/197620
[2023-03-27 22:26:36.668232] INFO: moduleinvoker: filter.v3 运行完成[5.38763s].
[2023-03-27 22:26:36.680712] INFO: moduleinvoker: features_short.v1 开始运行..
[2023-03-27 22:26:36.687270] INFO: moduleinvoker: 命中缓存
[2023-03-27 22:26:36.688795] INFO: moduleinvoker: features_short.v1 运行完成[0.008097s].
[2023-03-27 22:26:36.697074] INFO: moduleinvoker: stock_ranker_train.v6 开始运行..
[2023-03-27 22:26:41.529949] INFO: StockRanker: prepare data: test ..
[2023-03-27 22:26:41.538366] ERROR: moduleinvoker: module name: cached, module version: v2, trackeback: UnboundLocalError: local variable 'feature_cols' referenced before assignment
[2023-03-27 22:26:41.541670] ERROR: moduleinvoker: module name: stock_ranker_train, module version: v6, trackeback: UnboundLocalError: local variable 'feature_cols' referenced before assignment

---------------------------------------------------------------------------
UnboundLocalError                         Traceback (most recent call last)
<ipython-input-5-217deba77daf> in <module>
    167 )
    168 
--> 169 m20 = M.stock_ranker_train.v6(
    170     training_ds=m12.data,
    171     features=m5.data_1,

UnboundLocalError: local variable 'feature_cols' referenced before assignment