forked from zhostev/sqlib
-
Notifications
You must be signed in to change notification settings - Fork 0
/
prefect_switchable_workflow.py
187 lines (145 loc) · 6.77 KB
/
prefect_switchable_workflow.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
# 导入所需模块
from prefect import task, flow
from prefect import get_run_logger
import yaml
import pandas as pd
import mlflow
import qlib
from qlib.utils import init_instance_by_config
from qlib.contrib.strategy import TopkDropoutStrategy
from qlib.backtest import backtest, executor
from qlib.contrib.evaluate import risk_analysis
from data_handler.alpha158 import Alpha158
from strategy.topk_dropout import TopkDropoutStrategy
from qlib.data.dataset.handler import DataHandlerLP
from qlib.contrib.report.analysis_position.report import _calculate_report_data
from database_utils.db_utils import save_to_db
from utils.switchable_decorator import second_order_decorator as sod
from datetime import datetime
from utils.calc_group_return import get_group_return
import sys
def load_config():
config_name = "config/workflow_config_lightgbm_Alpha158_prefect_switchable.yaml"
with open(config_name, "r") as f:
config = yaml.safe_load(f)
return config
IS_DEBUGGING = sys.gettrace()
CFG = load_config()
USE_PREFECT = True if IS_DEBUGGING else CFG["use_prefect"]
EXP_NAME = CFG['handler_name'] + '_' + str(datetime.now()).split('.')[0].replace(':', '').replace('-', '').replace(' ', '_')
@sod(task, enabled=USE_PREFECT, name="model_data_init")
def model_data_init(config):
# Initialize QLib
use_prefect = config["use_prefect"]
provider_uri = config["qlib_init"]["provider_uri"]
region = config["qlib_init"]["region"]
qlib.init_qlib(provider_uri=provider_uri, region=region)
if use_prefect:
logger = get_run_logger()
logger.info("QLib initialized successfully")
# Initialize model
model_config = config["task"]["model"]
model = init_instance_by_config(model_config)
if use_prefect:
logger.info(f"Model initialized: {model}")
# Initialize data
# data_handler_config = config["task"]["dataset"]["kwargs"]["handler"]
# data_loader = QlibDataLoader(config=config['data_loader']['kwargs']['config'])
# df = data_loader.load(instruments='csi300', start_time='2010-01-01', end_time='2017-12-31')
data_handler_config = config["data_handler_config"]
handler_class = Alpha158 # if config['handler_name'] == 'Alpha158' else AlphaCustom
hd = handler_class(**data_handler_config)
dataset_conf = config["task"]["dataset"]
dataset_conf["kwargs"]["handler"] = hd
dataset = init_instance_by_config(dataset_conf)
if use_prefect:
logger.info(f"Dataset initialized: {dataset}")
# Reweighter = task_config.get("reweighter", None)
history = hd.fetch()
history = history.reset_index()
history.head()
save_to_db("history.db", "history_" + EXP_NAME, history)
return model, dataset
@sod(task, enabled=USE_PREFECT, name="train_and_predict")
def train_and_predict(model, dataset):
model.fit(dataset)
pred = model.predict(dataset)
if isinstance(pred, pd.Series):
pred = pred.to_frame("score")
pred["date"] = pred.index.get_level_values("datetime")
params = dict(segments="test", col_set="label", data_key=DataHandlerLP.DK_R)
label = dataset.prepare(**params)
save_to_db("pred.db", "pred_" + EXP_NAME, pred)
return pred, label
@sod(task, enabled=USE_PREFECT, name="strategy")
def strategy_simulator(config, pred):
STRATEGY_CONFIG = config["port_analysis_config"]["strategy"]["kwargs"]
STRATEGY_CONFIG["signal"] = pred
strategy_obj = TopkDropoutStrategy(**STRATEGY_CONFIG)
EXECUTOR_CONFIG = config["port_analysis_config"]["executor"]["kwargs"]
executor_obj = executor.SimulatorExecutor(**EXECUTOR_CONFIG)
return strategy_obj, executor_obj
@sod(task, enabled=USE_PREFECT, name="backtest_record")
def backtest_record(config, strategy_obj, executor_obj):
backtest_config = config["port_analysis_config"]["backtest"]
portfolio_metric_dict, indicator_dict = backtest(
executor=executor_obj, strategy=strategy_obj, **backtest_config
)
FREQ = "day"
analysis_freq = "{0}{1}".format(*qlib.utils.time.Freq.parse(FREQ))
portfolio_metrics = portfolio_metric_dict.get(analysis_freq)
report_normal = portfolio_metrics[0]
positions_normal = portfolio_metrics[1]
# convert report_normal to DataFrame and store it in database
report_df = report_normal.copy()
cumreport_df = _calculate_report_data(report_df)
cumreport_df = cumreport_df.iloc[1:]
cumreport_df.index = pd.to_datetime(cumreport_df.index)
cumreport_df["date"] = cumreport_df.index
save_to_db("report_normal.db", "report_normal_" + EXP_NAME, cumreport_df)
# get indicators_normal
indicators_normal = indicator_dict.get(analysis_freq)[0]
indicators_df = indicators_normal.copy()
# indicators_df = _calculate_report_data(indicators_df)
indicators_df = indicators_df.iloc[1:]
indicators_df.index = pd.to_datetime(indicators_df.index)
indicators_df["date"] = indicators_df.index
save_to_db("indicators_normal.db", "indicators_" + EXP_NAME, indicators_df)
# return results
return report_df, indicators_normal
@sod(task, enabled=USE_PREFECT, name="risk_analysis")
def risk_analysis_(report_normal):
analysis = dict()
analysis["excess_return_without_cost"] = risk_analysis(
report_normal["return"] - report_normal["bench"]
)
analysis["excess_return_with_cost"] = risk_analysis(
report_normal["return"] - report_normal["bench"] - report_normal["cost"]
)
analysis_df = pd.concat(analysis)
save_to_db("analysis_df.db", "analysis_" + EXP_NAME, analysis_df)
return analysis_df
@sod(task, enabled=USE_PREFECT, name="group_return")
def group_return(config, pred_df: pd.DataFrame = None, label_df: pd.DataFrame = None, reverse: bool = False, N: int = 5, **kwargs):
N = config.get("n_group", N)
pred_label = pd.concat([label_df, pred_df], axis=1, sort=True)
pred_label.columns.values[0] = 'label'
kwargs['rangebreaks'] = config.get('rangebreaks')
group_return_df = get_group_return(pred_label, reverse, N, return_df=True, **kwargs)
save_to_db("group_return.db", "group_return_" + EXP_NAME, group_return_df)
return group_return_df
# 定义流程
@sod(flow, enabled=USE_PREFECT, name="qlib_workflow", description="Demo Prefect")
def run_workflow(config=CFG, name="qlib_workflow"):
mlflow.set_tracking_uri("https://localhost:5000")
mlflow.set_experiment("Cyberquant")
with mlflow.start_run() as run:
mlflow.lightgbm.autolog()
model, dataset = model_data_init(config)
pred, label = train_and_predict(model, dataset)
strategy_obj, executor_obj = strategy_simulator(config, pred)
report_df, indicators_normal = backtest_record(config, strategy_obj, executor_obj)
analysis_df = risk_analysis_(report_df)
group_return_df = group_return(config, pred, label)
if __name__ == "__main__":
run_workflow()