qlib实践3-测试用例-测试基础思路

获取数据
from qlib.data import D

从D中筛选出数据和指标

>> from qlib.data import D
>> from qlib.data.filter import NameDFilter, ExpressionDFilter
>> nameDFilter = NameDFilter(name_rule_re='SH[0-9]{4}55')
>> expressionDFilter = ExpressionDFilter(rule_expression='$close>Ref($close,1)')
>> instruments = D.instruments(market='csi300', filter_pipe=[nameDFilter, expressionDFilter])
>> fields = ['$close', '$volume', 'Ref($close, 1)', 'Mean($close, 3)', '$high-$low']
>> D.features(instruments, fields, start_time='2010-01-01', end_time='2017-12-31', freq='day').head().to_string()
'                              $close        $volume  Ref($close, 1)  Mean($close, 3)  $high-$low
... instrument  datetime
... SH600655    2010-01-04  2699.567383  158193.328125     2619.070312      2626.097738  124.580566
...             2010-01-08  2612.359619   77501.406250     2584.567627      2623.220133   83.373047
...             2010-01-11  2712.982422  160852.390625     2612.359619      2636.636556  146.621582
...             2010-01-12  2788.688232  164587.937500     2712.982422      2704.676758  128.413818
...             2010-01-13  2790.604004  145460.453125     2788.688232      2764.091553  128.413818'

>> from qlib.data import D
>> data = D.features(["sh600519"], ["(($high / $close) + ($open / $close)) * (($high / $close) + ($open / $close)) / (($high / $close) + ($open / $close))"], start_time="20200101")

>> from qlib.data.ops import *
>> f1 = Feature("high") / Feature("close")
>> f2 = Feature("open") / Feature("close")
>> f3 = f1 + f2
>> f4 = f3 * f3 / f3

>> data = D.features(["sh600519"], [f4], start_time="20200101")
>> data.head()

参考文档：Data Retrieval — QLib 0.8.6.99 documentation

================================================================

workflow_by_code用例

1. 加载数据

qlib.init(provider_uri=provider_uri, region=REG_CN)

2. 设置模型数据范围和学习时间范围

market = "csi300"

data_handler_config = {
    "start_time": "2021-01-01",
    "end_time": "2022-12-31",
    "fit_start_time": "2021-01-01",
    "fit_end_time": "2021-12-31",
    "instruments": market,
}

3. 设置模型及参数

task = {
    "model": {
        "class": "LGBModel",
        "module_path": "qlib.contrib.model.gbdt",
        "kwargs": {
            "loss": "mse",
            "colsample_bytree": 0.8879,
            "learning_rate": 0.0421,
            "subsample": 0.8789,
            "lambda_l1": 205.6999,
            "lambda_l2": 580.9768,
            "max_depth": 8,
            "num_leaves": 210,
            "num_threads": 20,
        },
    },

4. 设置数据集和数据集处理范围

"dataset": {
        "class": "DatasetH",
        "module_path": "qlib.data.dataset",
        "kwargs": {
            "handler": {
                "class": "Alpha158",
                "module_path": "qlib.contrib.data.handler",
                "kwargs": data_handler_config, #步骤1中的时间范围和股票集合
            },
            "segments": {
                "train": ("2021-01-01", "2021-12-31"),
                "valid": ("2021-07-01", "2021-12-31"),
                "test": ("2022-01-01", "2022-12-01"),
            },
        },
    },

5. 模型初始化

model = init_instance_by_config(task["model"])
dataset = init_instance_by_config(task["dataset"])

6. 模型学习

# start exp to train model
with R.start(experiment_name="train_model"):
    R.log_params(**flatten_dict(task))

    model.fit(dataset)

    R.save_objects(trained_model=model)
    rid = R.get_recorder().id

7. 预测、回测、分析

# backtest and analysis
with R.start(experiment_name="backtest_analysis"):
    recorder = R.get_recorder(recorder_id=rid, experiment_name="train_model")
    model = recorder.load_object("trained_model")

    # prediction
    recorder = R.get_recorder()
    ba_rid = recorder.id
    sr = SignalRecord(model, dataset, recorder)
    sr.generate()

    # backtest & analysis
    par = PortAnaRecord(recorder, port_analysis_config, "day")
    par.generate()

7.1.其中PortAnaRecord做回测和分析，用到了一组新参数

port_analysis_config = {
    "executor": {
        #交易执行agent
        "class": "SimulatorExecutor",
        "module_path": "qlib.backtest.executor",
        "kwargs": {
            "time_per_step": "day",# 回测频率
            "generate_portfolio_metrics": True,
        },
    },
    "strategy": {
        "class": "TopkDropoutStrategy",
        "module_path": "qlib.contrib.strategy.signal_strategy",
        "kwargs": {
            "model": model,
            "dataset": dataset,
            "topk": 50,
            "n_drop": 5,
        },
    },
    "backtest": {
        "start_time": "2022-01-01",
        "end_time": "2022-12-01",
        "account": 100000000,
        "benchmark": benchmark, #基准收益，默认是沪深300， SH000300
        "exchange_kwargs": {
            "freq": "day",
            "limit_threshold": 0.095, #涨跌停设定 
            "deal_price": "close",
            "open_cost": 0.0005, #手续费设置
            "close_cost": 0.0015,
            "min_cost": 5,
        },
    },
}

可以这么理解，先对数据进行建立LGB模型，然后对数据进行预测，对预测的结果使用TopkDropout策略进行回测执行。

8. 图表分析

其中该模型的图表分析直接加载了磁盘中的pkl文件，后续将自行再测试一下

from qlib.contrib.report import analysis_model, analysis_position
from qlib.data import D
#模型加载，从experiment中加载
recorder = R.get_recorder(recorder_id=ba_rid, experiment_name="backtest_analysis")
pred_df = recorder.load_object("pred.pkl")

#从磁盘中加载3个不同的pkl，与回测分析进行对比？
report_normal_df = recorder.load_object("portfolio_analysis/report_normal_1day.pkl")
positions = recorder.load_object("portfolio_analysis/positions_normal_1day.pkl")
analysis_df = recorder.load_object("portfolio_analysis/port_analysis_1day.pkl")


#绘图
analysis_position.report_graph(report_normal_df)

图表指标，

cum bench: 基准指数的累计收益

cum return wo cost: 不计交易成本的策略累计收益

cum return w cost: 计入交易成本的策略累计收益

return w cost mdd: 计入交易成本的收益最大回撤

cum ex return wo cost: 不计交易成本的策略累计超额收益

turnover：策略换手率