liuyu/qlib
1
0
Fork 0
mirror of https://github.com/microsoft/qlib.git synced 2026-07-02 18:40:58 +08:00
Code Activity

fix comments & add VAStrategy & add trade indicator

Browse Source
This commit is contained in:
bxdd
2021-06-14 21:32:18 +08:00
parent 46d253b457 40416d8c30
commit f78e90171b
31 changed files with 1536 additions and 385 deletions
Download Patch File Download Diff File Expand all files Collapse all files

52
examples/benchmarks/TCTS/TCTS.md Normal file
View File

@@ -0,0 +1,52 @@
# Temporally Correlated Task Scheduling for Sequence Learning
We provide the [code](https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/pytorch_tcts.py) for reproducing the stock trend forecasting experiments.
### Background
Sequence learning has attracted much research attention from the machine learning community in recent years. In many applications, a sequence learning task is usually associated with multiple temporally correlated auxiliary tasks, which are different in terms of how much input information to use or which future step to predict. In stock trend forecasting, as demonstrated in Figure1, one can predict the price of a stock in different future days (e.g., tomorrow, the day after tomorrow). In this paper, we propose a framework to make use of those temporally correlated tasks to help each other.
<p align="center">
<img src="task_description.png" width="600" height="200"/>
</p>
### Method
Given that there are usually multiple temporally correlated tasks, the key challenge lies in which tasks to use and when to use them in the training process. In this work, we introduce a learnable task scheduler for sequence learning, which adaptively selects temporally correlated tasks during the training process. The scheduler accesses the model status and the current training data (e.g., in current minibatch), and selects the best auxiliary task to help the training of the main task. The scheduler and the model for the main task are jointly trained through bi-level optimization: the scheduler is trained to maximize the validation performance of the model, and the model is trained to minimize the training loss guided by the scheduler. The process is demonstrated in Figure2.
<p align="center">
<img src="workflow.png"/>
</p>
At step <img src="https://render.githubusercontent.com/render/math?math=s">, with training data <img src="https://render.githubusercontent.com/render/math?math=x_s,y_s">, the scheduler <img src="https://render.githubusercontent.com/render/math?math=\varphi"> chooses a suitable task <img src="https://render.githubusercontent.com/render/math?math=T_{i_s}"> (green solid lines) to update the model <img src="https://render.githubusercontent.com/render/math?math=f"> (blue solid lines). After <img src="https://render.githubusercontent.com/render/math?math=S"> steps, we evaluate the model <img src="https://render.githubusercontent.com/render/math?math=f"> on the validation set and update the scheduler <img src="https://render.githubusercontent.com/render/math?math=\varphi"> (green dashed lines).
### DataSet
* We use the historical transaction data for 300 stocks on [CSI300](http://www.csindex.com.cn/en/indices/index-detail/000300) from 01/01/2008 to 08/01/2020.
* We split the data into training (01/01/2008-12/31/2013), validation (01/01/2014-12/31/2015), and test sets (01/01/2016-08/01/2020) based on the transaction time.
### Experiments
#### Task Description
* The main tasks <img src="https://render.githubusercontent.com/render/math?math=T_k"> (<img src="https://render.githubusercontent.com/render/math?math=task_k"> in Figure1) refers to forecasting return of stock <img src="https://render.githubusercontent.com/render/math?math=i"> as following,
<div align=center>
<img src="https://render.githubusercontent.com/render/math?math=r_{i}^k = \frac{\price_i^{t+k}}{\price_i^{t+k-1}} - 1">
</div>
* Temporally correlated task sets <img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_k = \{T_1, T_2, ... , T_k\}">, in this paper, <img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_3">, <img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_5"> and <img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_10"> are used.
#### Baselines
* GRU/MLP/LightGBM (LGB)/Graph Attention Networks (GAT)
* Multi-task learning (MTL): In multi-task learning, multiple tasks are jointly trained and mutually boosted. Each task is treated equally, while in our setting, we focus on the main task.
* Curriculum transfer learning (CL): Transfer learning also leverages auxiliary tasks to boost the main task. [Curriculum transfer learning](https://arxiv.org/pdf/1804.00810.pdf) is one kind of transfer learning which schedules auxiliary tasks according to certain rules. Our problem can also be regarded as a special kind of transfer learning, where the auxiliary tasks are temporally correlated with the main task. Our learning process is dynamically controlled by a scheduler rather than some pre-defined rules. In the CL baseline, we start from the task <img src="https://render.githubusercontent.com/render/math?math=T_1" >, then <img src="https://render.githubusercontent.com/render/math?math=T_2" >, and gradually move to the last one.
#### Result
| Methods | <img src="https://render.githubusercontent.com/render/math?math=T_1" > | <img src="https://render.githubusercontent.com/render/math?math=T_2"> | <img src="https://render.githubusercontent.com/render/math?math=T_3"> |
| :----: | :----: | :----: | :----: |
| GRU | 0.049 / 1.903 | 0.018 / 1.972 | 0.014 / 1.989 |
| MLP | 0.023 / 1.961 | 0.022 / 1.962 | 0.015 / 1.978 |
| LGB | 0.038 / 1.883 | 0.023 / 1.952 | 0.007 / 1.987 |
| GAT | 0.052 / 1.898 | 0.024 / 1.954 | 0.015 / 1.973 |
| MTL(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_3">) | 0.061 / 1.862 | 0.023 / 1.942 | 0.012 / 1.956 |
| CL(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_3">) | 0.051 / 1.880 | 0.028 / 1.941 | 0.016 / 1.962 |
| Ours(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_3">) | 0.071 / 1.851 | 0.030 / 1.939 | 0.017 / 1.963 |
| MTL(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_5">) | 0.057 / 1.875 | 0.021 / 1.939 | 0.017 / 1.959 |
| CL(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_5">) | 0.056 / 1.877 | 0.028 / 1.942 | 0.015 / 1.962 |
| Ours(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_5">) | 0.075 / 1.849 | 0.032 /1.939 | 0.021 / 1.955 |
| MTL(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_{10}">) | 0.052 / 1.882 | 0.020 / 1.947 | 0.019 / 1.952 |
| CL(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_{10}">) | 0.051 / 1.882 | 0.028 / 1.950 | 0.016 / 1.961 |
| Ours(<img src="https://render.githubusercontent.com/render/math?math=\mathcal{T}_{10}">) | 0.067 / 1.867 | 0.030 / 1.960 | 0.022 / 1.942|

BIN
examples/benchmarks/TCTS/task_description.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 25 KiB

BIN
examples/benchmarks/TCTS/workflow.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 29 KiB

93
examples/benchmarks/TCTS/workflow_config_tcts_Alpha360.yaml Normal file
View File

@@ -0,0 +1,93 @@
qlib_init:
provider_uri: "~/.qlib/qlib_data/cn_data"
region: cn
market: &market csi300
benchmark: &benchmark SH000300
data_handler_config: &data_handler_config
start_time: 2008-01-01
end_time: 2020-08-01
fit_start_time: 2008-01-01
fit_end_time: 2014-12-31
instruments: *market
infer_processors:
- class: RobustZScoreNorm
kwargs:
fields_group: feature
clip_outlier: true
- class: Fillna
kwargs:
fields_group: feature
learn_processors:
- class: DropnaLabel
- class: CSRankNorm
kwargs:
fields_group: label
label: ["Ref($close, -2) / Ref($close, -1) - 1",
"Ref($close, -3) / Ref($close, -1) - 1",
"Ref($close, -4) / Ref($close, -1) - 1",
"Ref($close, -5) / Ref($close, -1) - 1",
"Ref($close, -6) / Ref($close, -1) - 1"]
port_analysis_config: &port_analysis_config
strategy:
class: TopkDropoutStrategy
module_path: qlib.contrib.strategy.strategy
kwargs:
topk: 50
n_drop: 5
backtest:
verbose: False
limit_threshold: 0.095
account: 100000000
benchmark: *benchmark
deal_price: close
open_cost: 0.0005
close_cost: 0.0015
min_cost: 5
task:
model:
class: TCTS
module_path: qlib.contrib.model.pytorch_tcts
kwargs:
d_feat: 6
hidden_size: 64
num_layers: 2
dropout: 0.0
n_epochs: 200
lr: 1e-3
early_stop: 20
batch_size: 800
metric: loss
loss: mse
GPU: 0
fore_optimizer: adam
weight_optimizer: adam
output_dim: 5
fore_lr: 5e-7
weight_lr: 5e-7
steps: 3
target_label: 0
dataset:
class: DatasetH
module_path: qlib.data.dataset
kwargs:
handler:
class: Alpha360
module_path: qlib.contrib.data.handler
kwargs: *data_handler_config
segments:
train: [2008-01-01, 2014-12-31]
valid: [2015-01-01, 2016-12-31]
test: [2017-01-01, 2020-08-01]
record:
- class: SignalRecord
module_path: qlib.workflow.record_temp
kwargs: {}
- class: SigAnaRecord
module_path: qlib.workflow.record_temp
kwargs:
ana_long_short: False
ann_scaler: 252
- class: PortAnaRecord
module_path: qlib.workflow.record_temp
kwargs:
config: *port_analysis_config

8
examples/model_rolling/task_manager_rolling.py
View File

@@ -4,6 +4,7 @@
"""
This example shows how a TrainerRM works based on TaskManager with rolling tasks.
After training, how to collect the rolling results will be shown in task_collecting.
Based on the ability of TaskManager, `worker` method offer a simple way for multiprocessing.
"""
from pprint import pprint
@@ -13,7 +14,7 @@ import qlib
from qlib.config import REG_CN
from qlib.workflow import R
from qlib.workflow.task.gen import RollingGen, task_generator
from qlib.workflow.task.manage import TaskManager
from qlib.workflow.task.manage import TaskManager, run_task
from qlib.workflow.task.collect import RecorderCollector
from qlib.model.ens.group import RollingGroup
from qlib.model.trainer import TrainerRM
@@ -68,6 +69,11 @@ class RollingTaskExample:
trainer = TrainerRM(self.experiment_name, self.task_pool)
trainer.train(tasks)
def worker(self):
# train tasks by other progress or machines for multiprocessing. It is same as TrainerRM.worker.
print("========== worker ==========")
run_task(task_train, self.task_pool, experiment_name=self.experiment_name)
def task_collecting(self):
print("========== task_collecting ==========")

2
examples/nested_decision_execution/README.md
View File

@@ -1,6 +1,6 @@
# Nested Decision Execution
This worflow is an example for nested decision execution in backtesting. Qlib supports nested decision execution in backtesting. It means that users can use different strategies to make trade decision in different frequencies.
This workflow is an example for nested decision execution in backtesting. Qlib supports nested decision execution in backtesting. It means that users can use different strategies to make trade decision in different frequencies.
## Weekly Portfolio Generation and Daily Order Execution

175
examples/nested_decision_execution/workflow.py
View File

@@ -19,10 +19,10 @@ class NestedDecisonExecutionWorkflow:
benchmark = "SH000300"
data_handler_config = {
"start_time": "2008-01-01",
"end_time": "2021-01-20",
"fit_start_time": "2008-01-01",
"fit_end_time": "2014-12-31",
"start_time": "2010-01-01",
"end_time": "2021-05-28",
"fit_start_time": "2010-01-01",
"fit_end_time": "2017-12-31",
"instruments": market,
}
@@ -52,9 +52,9 @@ class NestedDecisonExecutionWorkflow:
"kwargs": data_handler_config,
},
"segments": {
"train": ("2008-01-01", "2014-12-31"),
"valid": ("2015-01-01", "2016-12-31"),
"test": ("2017-01-01", "2021-01-20"),
"train": ("2010-01-01", "2017-12-31"),
"valid": ("2018-01-01", "2019-12-31"),
"test": ("2020-01-01", "2021-05-28"),
},
},
},
@@ -67,33 +67,45 @@ class NestedDecisonExecutionWorkflow:
"kwargs": {
"time_per_step": "week",
"inner_executor": {
"class": "SimulatorExecutor",
"class": "NestedExecutor",
"module_path": "qlib.backtest.executor",
"kwargs": {
"time_per_step": "day",
"verbose": True,
"generate_report": True,
"inner_executor": {
"class": "SimulatorExecutor",
"module_path": "qlib.backtest.executor",
"kwargs": {
"time_per_step": "15min",
"generate_report": True,
"verbose": True,
},
},
"inner_strategy": {
"class": "TWAPStrategy",
"module_path": "qlib.contrib.strategy.rule_strategy",
},
"show_indicator": True,
},
},
"inner_strategy": {
"class": "SBBStrategyEMA",
"class": "VAStrategy",
"module_path": "qlib.contrib.strategy.rule_strategy",
"kwargs": {
"freq": "day",
"instruments": market,
},
},
"generate_report": True,
"track_data": True,
"show_indicator": True,
},
},
"backtest": {
"start_time": "2017-01-01",
"end_time": "2020-08-01",
"start_time": "2020-09-20",
"end_time": "2021-05-28",
"account": 100000000,
"benchmark": benchmark,
"exchange_kwargs": {
"freq": "day",
"freq": "1min",
"limit_threshold": 0.095,
"deal_price": "close",
"open_cost": 0.0005,
@@ -105,11 +117,40 @@ class NestedDecisonExecutionWorkflow:
def _init_qlib(self):
"""initialize qlib"""
provider_uri = "~/.qlib/qlib_data/cn_data" # target_dir
if not exists_qlib_data(provider_uri):
print(f"Qlib data is not found in {provider_uri}")
GetData().qlib_data(target_dir=provider_uri, region=REG_CN)
qlib.init(provider_uri=provider_uri, region=REG_CN)
provider_uri_day = "/data1/v-xiabi/qlib/qlib_data/cn_data" # target_dir
GetData().qlib_data(target_dir=provider_uri_day, region=REG_CN, version="v2", exists_skip=True)
# provider_uri_1min = HIGH_FREQ_CONFIG.get("provider_uri")
provider_uri_1min = "/data1/v-xiabi/qlib/qlib_data/cn_data_highfreq"
GetData().qlib_data(
target_dir=provider_uri_1min, interval="1min", region=REG_CN, version="v2", exists_skip=True
)
provider_uri_map = {"1min": provider_uri_1min, "day": provider_uri_day}
client_config = {
"calendar_provider": {
"class": "LocalCalendarProvider",
"module_path": "qlib.data.data",
"kwargs": {
"backend": {
"class": "FileCalendarStorage",
"module_path": "qlib.data.storage.file_storage",
"kwargs": {"provider_uri_map": provider_uri_map},
}
},
},
"feature_provider": {
"class": "LocalFeatureProvider",
"module_path": "qlib.data.data",
"kwargs": {
"backend": {
"class": "FileFeatureStorage",
"module_path": "qlib.data.storage.file_storage",
"kwargs": {"provider_uri_map": provider_uri_map},
}
},
},
}
qlib.init(provider_uri=provider_uri_day, **client_config)
def _train_model(self, model, dataset):
with R.start(experiment_name="train"):
@@ -141,7 +182,7 @@ class NestedDecisonExecutionWorkflow:
with R.start(experiment_name="backtest"):
recorder = R.get_recorder()
par = PortAnaRecord(recorder, self.port_analysis_config, "day")
par = PortAnaRecord(recorder, self.port_analysis_config, "15minute")
par.generate()
def collect_data(self):
@@ -165,98 +206,6 @@ class NestedDecisonExecutionWorkflow:
for trade_decision in data_generator:
print(trade_decision)
def _init_qlib_with_backend(self):
provider_uri_1min = HIGH_FREQ_CONFIG.get("provider_uri")
if not exists_qlib_data(provider_uri_1min):
print(f"Qlib data is not found in {provider_uri_1min}")
GetData().qlib_data(target_dir=provider_uri_1min, interval="1min", region=REG_CN)
# TODO: update latest data
provider_uri_day = "~/.qlib/qlib_data/cn_data" # target_dir
if not exists_qlib_data(provider_uri_day):
print(f"Qlib data is not found in {provider_uri_day}")
GetData().qlib_data(target_dir=provider_uri_day, region=REG_CN)
provider_uri_map = {"1min": provider_uri_1min, "day": provider_uri_day}
client_config = {
"calendar_provider": {
"class": "LocalCalendarProvider",
"module_path": "qlib.data.data",
"kwargs": {
"backend": {
"class": "FileCalendarStorage",
"module_path": "qlib.data.storage.file_storage",
"kwargs": {"provider_uri_map": provider_uri_map},
}
},
},
"feature_provider": {
"class": "LocalFeatureProvider",
"module_path": "qlib.data.data",
"kwargs": {
"backend": {
"class": "FileFeatureStorage",
"module_path": "qlib.data.storage.file_storage",
"kwargs": {"provider_uri_map": provider_uri_map},
}
},
},
}
qlib.init(provider_uri=provider_uri_day, **client_config)
def _get_highfreq_config(self, model, dataset):
executor_config = self.port_analysis_config["executor"]
# update executor with hierarchical decison freq ["day", "1min"]
executor_config["kwargs"]["time_per_step"] = "day"
executor_config["kwargs"]["inner_executor"]["kwargs"]["time_per_step"] = "15min"
backtest_config = self.port_analysis_config["backtest"]
# yahoo highfreq data time
backtest_config["start_time"] = "2020-09-20"
backtest_config["end_time"] = "2021-01-20"
# update benchmark, yahoo data don't have SH000300
instruments = D.instruments(market="csi300")
instrument_list = D.list_instruments(instruments=instruments, as_list=True)
backtest_config["benchmark"] = instrument_list
# update exchange config
backtest_config["exchange_kwargs"]["freq"] = "1min"
# set strategy
strategy_config = {
"class": "TopkDropoutStrategy",
"module_path": "qlib.contrib.strategy.model_strategy",
"kwargs": {
"model": model,
"dataset": dataset,
"topk": 50,
"n_drop": 5,
},
}
return executor_config, strategy_config, backtest_config
def backtest_highfreq(self):
self._init_qlib_with_backend()
model = init_instance_by_config(self.task["model"])
dataset = init_instance_by_config(self.task["dataset"])
self._train_model(model, dataset)
executor_config, strategy_config, backtest_config = self._get_highfreq_config(model, dataset)
highfreq_port_analysis_config = {
"executor": executor_config,
"strategy": strategy_config,
"backtest": backtest_config,
}
with R.start(experiment_name="backtest_highfreq"):
recorder = R.get_recorder()
par = PortAnaRecord(recorder, highfreq_port_analysis_config, "day")
par.generate()
if __name__ == "__main__":
fire.Fire(NestedDecisonExecutionWorkflow)

20
examples/online_srv/online_management_simulate.py
View File

@@ -5,6 +5,7 @@
This example is about how can simulate the OnlineManager based on rolling tasks.
"""
from pprint import pprint
import fire
import qlib
from qlib.model.trainer import DelayTrainerR, DelayTrainerRM, TrainerR, TrainerRM
@@ -13,7 +14,7 @@ from qlib.workflow.online.manager import OnlineManager
from qlib.workflow.online.strategy import RollingStrategy
from qlib.workflow.task.gen import RollingGen
from qlib.workflow.task.manage import TaskManager
from qlib.tests.config import CSI100_RECORD_LGB_TASK_CONFIG, CSI100_RECORD_XGBOOST_TASK_CONFIG
from qlib.tests.config import CSI100_RECORD_LGB_TASK_CONFIG_ONLINE, CSI100_RECORD_XGBOOST_TASK_CONFIG_ONLINE
class OnlineSimulationExample:
@@ -22,8 +23,8 @@ class OnlineSimulationExample:
provider_uri="~/.qlib/qlib_data/cn_data",
region="cn",
exp_name="rolling_exp",
task_url="mongodb://10.0.0.4:27017/",
task_db_name="rolling_db",
task_url="mongodb://10.0.0.4:27017/", # not necessary when using TrainerR or DelayTrainerR
task_db_name="rolling_db", # not necessary when using TrainerR or DelayTrainerR
task_pool="rolling_task",
rolling_step=80,
start_time="2018-09-10",
@@ -46,7 +47,7 @@ class OnlineSimulationExample:
tasks (dict or list[dict]): a set of the task config waiting for rolling and training
"""
if tasks is None:
tasks = [CSI100_RECORD_XGBOOST_TASK_CONFIG, CSI100_RECORD_LGB_TASK_CONFIG]
tasks = [CSI100_RECORD_XGBOOST_TASK_CONFIG_ONLINE, CSI100_RECORD_LGB_TASK_CONFIG_ONLINE]
self.exp_name = exp_name
self.task_pool = task_pool
self.start_time = start_time
@@ -59,7 +60,7 @@ class OnlineSimulationExample:
self.rolling_gen = RollingGen(
step=rolling_step, rtype=RollingGen.ROLL_SD, ds_extra_mod_func=None
) # The rolling tasks generator, ds_extra_mod_func is None because we just need to simulate to 2018-10-31 and needn't change the handler end time.
self.trainer = DelayTrainerRM(self.exp_name, self.task_pool) # Also can be TrainerR, TrainerRM, DelayTrainerR
self.trainer = TrainerRM(self.exp_name, self.task_pool) # Also can be TrainerR, TrainerRM, DelayTrainerR
self.rolling_online_manager = OnlineManager(
RollingStrategy(exp_name, task_template=tasks, rolling_gen=self.rolling_gen),
trainer=self.trainer,
@@ -85,6 +86,15 @@ class OnlineSimulationExample:
print("========== signals ==========")
print(self.rolling_online_manager.get_signals())
def worker(self):
# train tasks by other progress or machines for multiprocessing
# FIXME: only can call after finishing simulation when using DelayTrainerRM, or there will be some exception.
print("========== worker ==========")
if isinstance(self.trainer, TrainerRM):
self.trainer.worker()
else:
print(f"{type(self.trainer)} is not supported for worker.")
if __name__ == "__main__":
## to run all workflow automatically with your own parameters, use the command below

28
examples/online_srv/rolling_online_management.py
View File

@@ -13,11 +13,13 @@ Finally, the OnlineManager will finish second routine and update all strategies.
import os
import fire
import qlib
from qlib.model.trainer import DelayTrainerR, DelayTrainerRM, TrainerR, TrainerRM, end_task_train, task_train
from qlib.workflow import R
from qlib.workflow.online.strategy import RollingStrategy
from qlib.workflow.task.gen import RollingGen
from qlib.workflow.online.manager import OnlineManager
from qlib.tests.config import CSI100_RECORD_XGBOOST_TASK_CONFIG, CSI100_RECORD_LGB_TASK_CONFIG
from qlib.tests.config import CSI100_RECORD_XGBOOST_TASK_CONFIG_ROLLING, CSI100_RECORD_LGB_TASK_CONFIG_ROLLING
from qlib.workflow.task.manage import TaskManager
class RollingOnlineExample:
@@ -25,16 +27,17 @@ class RollingOnlineExample:
self,
provider_uri="~/.qlib/qlib_data/cn_data",
region="cn",
task_url="mongodb://10.0.0.4:27017/",
task_db_name="rolling_db",
trainer=DelayTrainerRM(), # you can choose from TrainerR, TrainerRM, DelayTrainerR, DelayTrainerRM
task_url="mongodb://10.0.0.4:27017/", # not necessary when using TrainerR or DelayTrainerR
task_db_name="rolling_db", # not necessary when using TrainerR or DelayTrainerR
rolling_step=550,
tasks=None,
add_tasks=None,
):
if add_tasks is None:
add_tasks = [CSI100_RECORD_LGB_TASK_CONFIG]
add_tasks = [CSI100_RECORD_LGB_TASK_CONFIG_ROLLING]
if tasks is None:
tasks = [CSI100_RECORD_XGBOOST_TASK_CONFIG]
tasks = [CSI100_RECORD_XGBOOST_TASK_CONFIG_ROLLING]
mongo_conf = {
"task_url": task_url, # your MongoDB url
"task_db_name": task_db_name, # database name
@@ -53,17 +56,28 @@ class RollingOnlineExample:
RollingGen(step=rolling_step, rtype=RollingGen.ROLL_SD),
)
)
self.rolling_online_manager = OnlineManager(strategies)
self.trainer = trainer
self.rolling_online_manager = OnlineManager(strategies, trainer=self.trainer)
_ROLLING_MANAGER_PATH = (
".RollingOnlineExample" # the OnlineManager will dump to this file, for it can be loaded when calling routine.
)
def worker(self):
# train tasks by other progress or machines for multiprocessing
print("========== worker ==========")
if isinstance(self.trainer, TrainerRM):
for task in self.tasks + self.add_tasks:
name_id = task["model"]["class"]
self.trainer.worker(experiment_name=name_id)
else:
print(f"{type(self.trainer)} is not supported for worker.")
# Reset all things to the first status, be careful to save important data
def reset(self):
for task in self.tasks + self.add_tasks:
name_id = task["model"]["class"]
TaskManager(task_pool=name_id).remove()
exp = R.get_exp(experiment_name=name_id)
for rid in exp.list_recorders():
exp.delete_recorder(rid)

12
examples/workflow_by_code.ipynb
View File

@@ -362,8 +362,9 @@
],
"metadata": {
"kernelspec": {
"name": "pythonjvsc74a57bd0fcc004278713aaede7c629a6a43738a929cb09abb52817d4f72eb70db44cd87b",
"display_name": "Python 3.8 ('qlib_backtest': conda)"
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
@@ -375,7 +376,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8"
"version": "3.8.3"
},
"toc": {
"base_numbering": 1,
@@ -389,11 +390,6 @@
"toc_position": {},
"toc_section_display": true,
"toc_window_display": false
},
"metadata": {
"interpreter": {
"hash": "fcc004278713aaede7c629a6a43738a929cb09abb52817d4f72eb70db44cd87b"
}
}
},
"nbformat": 4,

8
qlib/backtest/__init__.py
View File

@@ -4,8 +4,8 @@
from .account import Account
from .exchange import Exchange
from .executor import BaseExecutor
from .backtest import backtest as backtest_func
from .backtest import collect_data as data_generator
from .backtest import backtest_loop
from .backtest import collect_data_loop
from .utils import CommonInfrastructure
from .order import Order
@@ -116,7 +116,7 @@ def backtest(start_time, end_time, strategy, executor, benchmark="SH000300", acc
trade_strategy, trade_executor = get_strategy_executor(
start_time, end_time, strategy, executor, benchmark, account, exchange_kwargs
)
report_dict = backtest_func(start_time, end_time, trade_strategy, trade_executor)
report_dict = backtest_loop(start_time, end_time, trade_strategy, trade_executor)
return report_dict
@@ -126,6 +126,6 @@ def collect_data(start_time, end_time, strategy, executor, benchmark="SH000300",
trade_strategy, trade_executor = get_strategy_executor(
start_time, end_time, strategy, executor, benchmark, account, exchange_kwargs
)
report_dict = yield from data_generator(start_time, end_time, trade_strategy, trade_executor)
report_dict = yield from collect_data_loop(start_time, end_time, trade_strategy, trade_executor)
return report_dict

3
qlib/backtest/account.py
View File

@@ -7,7 +7,7 @@ import warnings
import pandas as pd
from .position import Position
from .report import Report
from .report import Report, Indicator
from .order import Order
@@ -42,6 +42,7 @@ class Account:
def reset_report(self, freq, benchmark_config):
self.report = Report(freq, benchmark_config)
self.indicator = Indicator()
self.positions = {}
self.rtn = 0
self.ct = 0

29
qlib/backtest/backtest.py
View File

@@ -2,8 +2,25 @@
# Licensed under the MIT License.
def backtest(start_time, end_time, trade_strategy, trade_executor):
def backtest_loop(start_time, end_time, trade_strategy, trade_executor):
"""backtest funciton for the interaction of the outermost strategy and executor in the nested decison execution
Parameters
----------
start_time : pd.Timestamp|str
closed start time for backtest
end_time : pd.Timestamp|str
closed end time for backtest
trade_strategy : BaseStrategy
the outermost portfolio strategy
trade_executor : BaseExecutor
the outermost executor
Returns
-------
report: Report
it records the trading report information
"""
trade_executor.reset(start_time=start_time, end_time=end_time)
level_infra = trade_executor.get_level_infra()
trade_strategy.reset(level_infra=level_infra)
@@ -16,8 +33,14 @@ def backtest(start_time, end_time, trade_strategy, trade_executor):
return trade_executor.get_report()
def collect_data(start_time, end_time, trade_strategy, trade_executor):
def collect_data_loop(start_time, end_time, trade_strategy, trade_executor):
"""Generator for collecting the trade decision data for rl training
Yields
-------
object
trade decision
"""
trade_executor.reset(start_time=start_time, end_time=end_time)
level_infra = trade_executor.get_level_infra()
trade_strategy.reset(level_infra=level_infra)
@@ -26,5 +49,3 @@ def collect_data(start_time, end_time, trade_strategy, trade_executor):
while not trade_executor.finished():
_trade_decision = trade_strategy.generate_trade_decision(_execute_result)
_execute_result = yield from trade_executor.collect_data(_trade_decision)
return trade_executor.get_report()

5
qlib/backtest/exchange.py
View File

@@ -342,7 +342,10 @@ class Exchange:
return -deal_amount
def generate_order_for_target_amount_position(self, target_position, current_position, start_time, end_time):
"""Parameter:
"""
Note: some future information is used in this function
Parameter:
target_position : dict { stock_id : amount }
current_postion : dict { stock_id : amount}
trade_unit : trade_unit

105
qlib/backtest/executor.py
View File

@@ -3,14 +3,14 @@ import warnings
import pandas as pd
from typing import Union
from ..utils import init_instance_by_config
from ..utils.resam import parse_freq
from .order import Order
from .exchange import Exchange
from .utils import TradeCalendarManager, CommonInfrastructure, LevelInfrastructure
from ..utils import init_instance_by_config
from ..utils.resam import parse_freq
from ..strategy.base import BaseStrategy
class BaseExecutor:
"""Base executor for trading"""
@@ -20,6 +20,7 @@ class BaseExecutor:
time_per_step: str,
start_time: Union[str, pd.Timestamp] = None,
end_time: Union[str, pd.Timestamp] = None,
show_indicator: bool = False,
generate_report: bool = False,
verbose: bool = False,
track_data: bool = False,
@@ -31,12 +32,14 @@ class BaseExecutor:
----------
time_per_step : str
trade time per trading step, used for genreate the trade calendar
show_indicator: bool, optional
whether to show indicators, such as FFR/PA/POS, .etc
generate_report : bool, optional
whether to generate report, by default False
verbose : bool, optional
whether to print trading info, by default False
track_data : bool, optional
whether to generate trade_decision, will be used when making data for multi-level training
whether to generate trade_decision, will be used when training rl agent
- If `self.track_data` is true, when making data for training, the input `trade_decision` of `execute` will be generated by `collect_data`
- Else, `trade_decision` will not be generated
common_infra : CommonInfrastructure, optional:
@@ -48,6 +51,7 @@ class BaseExecutor:
"""
self.time_per_step = time_per_step
self.show_indicator = show_indicator
self.generate_report = generate_report
self.verbose = verbose
self.track_data = track_data
@@ -103,11 +107,27 @@ class BaseExecutor:
Returns
----------
execute_result : List[object]
the executed result for trade decison
the executed result for trade decision
"""
raise NotImplementedError("execute is not implemented!")
def collect_data(self, trade_decision):
"""Generator for collecting the trade decision data for rl training
Parameters
----------
trade_decision : object
Returns
----------
execute_result : List[object]
the executed result for trade decision
Yields
-------
object
trade decision
"""
if self.track_data:
yield trade_decision
return self.execute(trade_decision)
@@ -122,6 +142,9 @@ class BaseExecutor:
"""Return all executors"""
return [self]
def get_trade_indicator(self):
return self.trade_account.indicator.trade_indicator
class NestedExecutor(BaseExecutor):
"""
@@ -129,8 +152,6 @@ class NestedExecutor(BaseExecutor):
- At each time `execute` is called, it will call the inner strategy and executor to execute the `trade_decision` in a higher frequency env.
"""
from ..strategy.base import BaseStrategy
def __init__(
self,
time_per_step: str,
@@ -138,6 +159,7 @@ class NestedExecutor(BaseExecutor):
inner_strategy: Union[BaseStrategy, dict],
start_time: Union[str, pd.Timestamp] = None,
end_time: Union[str, pd.Timestamp] = None,
show_indicator: bool = False,
generate_report: bool = False,
verbose: bool = False,
track_data: bool = False,
@@ -161,13 +183,14 @@ class NestedExecutor(BaseExecutor):
inner_executor, common_infra=common_infra, accept_types=BaseExecutor
)
self.inner_strategy = init_instance_by_config(
inner_strategy, common_infra=common_infra, accept_types=self.BaseStrategy
inner_strategy, common_infra=common_infra, accept_types=BaseStrategy
)
super(NestedExecutor, self).__init__(
time_per_step=time_per_step,
start_time=start_time,
end_time=end_time,
show_indicator=show_indicator,
generate_report=generate_report,
verbose=verbose,
track_data=track_data,
@@ -199,7 +222,7 @@ class NestedExecutor(BaseExecutor):
sub_level_infra = self.inner_executor.get_level_infra()
self.inner_strategy.reset(level_infra=sub_level_infra, outer_trade_decision=trade_decision)
def _update_trade_account(self):
def _update_trade_account(self, inner_indicators):
trade_step = self.trade_calendar.get_trade_step()
trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
self.trade_account.update_bar_count()
@@ -210,33 +233,44 @@ class NestedExecutor(BaseExecutor):
trade_exchange=self.trade_exchange,
)
self.trade_account.indicator.clear()
self.trade_account.indicator.agg_report_info(inner_indicators=inner_indicators)
self.trade_account.indicator.agg_FFR()
self.trade_account.indicator.agg_PA(inner_indicators=inner_indicators)
if self.show_indicator:
FFR_value = self.trade_account.indicator.get_statistics_FFR(method="value_weighted")
PA_value = self.trade_account.indicator.get_statistics_PA(method="value_weighted")
POS_values = self.trade_account.indicator.get_statistics_POS()
print(
"[Indicator({}) {:%Y-%m-%d}]: FFR: {}, PA: {}, POS: {}".format(
self.time_per_step, trade_start_time, FFR_value, PA_value, POS_values
)
)
def execute(self, trade_decision):
self._init_sub_trading(trade_decision)
execute_result = []
_inner_execute_result = None
while not self.inner_executor.finished():
_inner_trade_decision = self.inner_strategy.generate_trade_decision(_inner_execute_result)
_inner_execute_result = self.inner_executor.execute(trade_decision=_inner_trade_decision)
execute_result.extend(_inner_execute_result)
if hasattr(self, "trade_account"):
self._update_trade_account()
self.trade_calendar.step()
return execute_result
for _data in self.collect_data(trade_decision):
pass
return self._execute_result
def collect_data(self, trade_decision):
if self.track_data:
yield trade_decision
self.trade_calendar.step()
self._init_sub_trading(trade_decision)
execute_result = []
inner_indicators = []
_inner_execute_result = None
while not self.inner_executor.finished():
_inner_trade_decision = self.inner_strategy.generate_trade_decision(_inner_execute_result)
_inner_execute_result = yield from self.inner_executor.collect_data(trade_decision=_inner_trade_decision)
execute_result.extend(_inner_execute_result)
if hasattr(self, "trade_account"):
self._update_trade_account()
inner_indicators.append(self.inner_executor.get_trade_indicator())
if hasattr(self, "trade_account"):
self._update_trade_account(inner_indicators=inner_indicators)
self.trade_calendar.step()
self._execute_result = execute_result
return execute_result
def get_report(self):
@@ -261,6 +295,7 @@ class SimulatorExecutor(BaseExecutor):
time_per_step: str,
start_time: Union[str, pd.Timestamp] = None,
end_time: Union[str, pd.Timestamp] = None,
show_indicator: bool = False,
generate_report: bool = False,
verbose: bool = False,
track_data: bool = False,
@@ -279,6 +314,7 @@ class SimulatorExecutor(BaseExecutor):
time_per_step=time_per_step,
start_time=start_time,
end_time=end_time,
show_indicator=show_indicator,
generate_report=generate_report,
verbose=verbose,
track_data=track_data,
@@ -337,7 +373,7 @@ class SimulatorExecutor(BaseExecutor):
else:
if self.verbose:
print("[W {:%Y-%m-%d}]: {} wrong.".format(trade_start_time, order.stock_id))
print("[W {:%Y-%m-%d %H:%M:%S}]: {} wrong.".format(trade_start_time, order.stock_id))
# do nothing
pass
@@ -349,6 +385,25 @@ class SimulatorExecutor(BaseExecutor):
trade_end_time=trade_end_time,
trade_exchange=self.trade_exchange,
)
self.trade_account.indicator.clear()
self.trade_account.indicator.update_trade_info(trade_info=execute_result)
self.trade_account.indicator.update_FFR()
self.trade_account.indicator.update_PA(
freq=self.time_per_step, trade_start_time=trade_start_time, trade_end_time=trade_end_time
)
self.trade_account.indicator.record(trade_start_time=trade_start_time)
if self.show_indicator:
FFR_value = self.trade_account.indicator.get_statistics_FFR(method="value_weighted")
PA_value = self.trade_account.indicator.get_statistics_PA(method="value_weighted")
POS_values = self.trade_account.indicator.get_statistics_POS()
print(
"[Indicator({}) {:%Y-%m-%d %H:%M:%S}]: FFR: {}, PA: {}, POS: {}".format(
self.time_per_step, trade_start_time, FFR_value, PA_value, POS_values
)
)
self.trade_calendar.step()
return execute_result

195
qlib/backtest/report.py
View File

@@ -7,10 +7,11 @@ from logging import warning
import pandas as pd
import pathlib
import warnings
from pandas.core import groupby
from pandas.core.frame import DataFrame
from ..utils.resam import parse_freq, resam_ts_data
from ..utils.resam import parse_freq, resam_ts_data, get_higher_freq_feature
from ..data import D
from ..tests.config import CSI300_BENCH
@@ -79,19 +80,7 @@ class Report:
raise ValueError("benchmark freq can't be None!")
_codes = benchmark if isinstance(benchmark, list) else [benchmark]
fields = ["$close/Ref($close,1)-1"]
try:
_temp_result = D.features(_codes, fields, start_time, end_time, freq=freq, disk_cache=1)
except (ValueError, KeyError):
_, norm_freq = parse_freq(freq)
if norm_freq in ["month", "week", "day"]:
try:
_temp_result = D.features(_codes, fields, start_time, end_time, freq="day", disk_cache=1)
except (ValueError, KeyError):
_temp_result = D.features(_codes, fields, start_time, end_time, freq="1min", disk_cache=1)
elif norm_freq == "minute":
_temp_result = D.features(_codes, fields, start_time, end_time, freq="1min", disk_cache=1)
else:
raise ValueError(f"benchmark freq {freq} is not supported")
_temp_result, _ = get_higher_freq_feature(_codes, fields, start_time, end_time, freq=freq)
if len(_temp_result) == 0:
raise ValueError(f"The benchmark {_codes} does not exist. Please provide the right benchmark")
return _temp_result.groupby(level="datetime")[_temp_result.columns.tolist()[0]].mean().fillna(0)
@@ -122,11 +111,11 @@ class Report:
turnover_rate=None,
cost_rate=None,
stock_value=None,
bench_value=None,
):
# check data
if None in [
trade_start_time,
trade_end_time,
account_value,
cash,
return_rate,
@@ -135,8 +124,14 @@ class Report:
stock_value,
]:
raise ValueError(
"None in [trade_start_time, trade_end_time, account_value, cash, return_rate, turnover_rate, cost_rate, stock_value]"
"None in [trade_start_time, account_value, cash, return_rate, turnover_rate, cost_rate, stock_value]"
)
if trade_end_time is None and bench_value is None:
raise ValueError("Both trade_end_time and bench_value is None, benchmark is not usable.")
elif bench_value is None:
bench_value = self._sample_benchmark(self.bench, trade_start_time, trade_end_time)
# update report data
self.accounts[trade_start_time] = account_value
self.returns[trade_start_time] = return_rate
@@ -144,7 +139,7 @@ class Report:
self.costs[trade_start_time] = cost_rate
self.values[trade_start_time] = stock_value
self.cashes[trade_start_time] = cash
self.benches[trade_start_time] = self._sample_benchmark(self.bench, trade_start_time, trade_end_time)
self.benches[trade_start_time] = bench_value
# update latest_report_date
self.latest_report_time = trade_start_time
# finish daily report update
@@ -178,14 +173,162 @@ class Report:
index = r.index
self.init_vars()
for trade_time in index:
for trade_start_time in index:
self.update_report_record(
trade_time=trade_time,
account_value=r.loc[trade_time]["account"],
cash=r.loc[trade_time]["cash"],
return_rate=r.loc[trade_time]["return"],
turnover_rate=r.loc[trade_time]["turnover"],
cost_rate=r.loc[trade_time]["cost"],
stock_value=r.loc[trade_time]["value"],
bench_value=r.loc[trade_time]["bench"],
trade_start_time=trade_start_time,
account_value=r.loc[trade_start_time]["account"],
cash=r.loc[trade_start_time]["cash"],
return_rate=r.loc[trade_start_time]["return"],
turnover_rate=r.loc[trade_start_time]["turnover"],
cost_rate=r.loc[trade_start_time]["cost"],
stock_value=r.loc[trade_start_time]["value"],
bench_value=r.loc[trade_start_time]["bench"],
)
class Indicator:
def __init__(self):
self.indicator_his = dict()
self.trade_indicator = dict()
def __getitem__(self, key):
return self.trade_indicator[key]
def __setitem__(self, key, value):
self.trade_indicator[key] = value
def __contains__(self, key):
return key in self.trade_indicator
def clear(self):
self.trade_indicator = dict()
def record(self, trade_start_time):
self.indicator_his[trade_start_time] = pd.DataFrame(self.trade_indicator)
def update_trade_info(self, trade_info: list):
amount = dict()
deal_amount = dict()
trade_price = dict()
trade_cost = dict()
for order, _trade_val, _trade_cost, _trade_price in trade_info:
amount[order.stock_id] = order.amount * (order.direction * 2 - 1)
deal_amount[order.stock_id] = order.deal_amount * (order.direction * 2 - 1)
trade_price[order.stock_id] = _trade_price
trade_cost[order.stock_id] = _trade_cost
self["amount"] = pd.Series(amount)
self["deal_amount"] = pd.Series(deal_amount)
self["trade_price"] = pd.Series(trade_price)
self["trade_cost"] = pd.Series(trade_cost)
def update_FFR(self):
self["fulfill_rate"] = self["deal_amount"] / self["amount"]
def update_PA(self, freq, trade_start_time, trade_end_time, base_price="twap"):
base_price = base_price.lower()
instruments = list(self["amount"].index)
if base_price == "twap":
# too slow
# price_info, _ = get_higher_freq_feature(instruments, fields=["$close"], start_time=trade_start_time, end_time=trade_end_time, freq=freq)
# price_info = price_info.astype(float)
# self["base_price"] = price_info["$close"].groupby(level="instrument").mean()
self["base_price"] = self["trade_price"]
elif base_price == "vwap":
# too slow
price_info, _ = get_higher_freq_feature(
instruments,
fields=["$close", "$volume"],
start_time=trade_start_time,
end_time=trade_end_time,
freq=freq,
)
price_info = price_info.astype(float)
self["base_price"] = price_info.groupby(level="instrument").apply(
lambda x: (x["$close"] * x["$volume"]).sum() / x["$volume"].sum()
)
self["volume"] = price_info["$volume"].groupby(level="instrument").sum()
else:
raise ValueError(f"base_price {base_price} is not supported!")
self["pa"] = (self["trade_price"] - self["base_price"]) / self["base_price"]
def agg_report_info(self, inner_indicators):
amount = pd.Series()
deal_amount = pd.Series()
trade_price = pd.Series()
trade_cost = pd.Series()
for inner_indicator in inner_indicators:
amount = amount.add(inner_indicator["amount"], fill_value=0)
deal_amount = deal_amount.add(inner_indicator["deal_amount"], fill_value=0)
trade_price = trade_price.add(inner_indicator["trade_price"] * inner_indicator["deal_amount"], fill_value=0)
trade_cost = trade_cost.add(inner_indicator["trade_cost"], fill_value=0)
self["amount"] = amount
self["deal_amount"] = deal_amount
trade_price /= self["deal_amount"]
self["trade_price"] = trade_price
self["trade_cost"] = trade_cost
def agg_FFR(self):
self["fulfill_rate"] = self["deal_amount"] / self["amount"]
def agg_PA(self, inner_indicators, base_price="twap"):
base_price = base_price.lower()
if base_price == "twap":
base_price = pd.Series()
price_count = pd.Series()
for inner_indicator in inner_indicators:
base_price = base_price.add(inner_indicator["base_price"], fill_value=0)
price_count = price_count.add(pd.Series(1, index=inner_indicator["base_price"].index), fill_value=0)
base_price /= price_count
self["base_price"] = base_price
elif base_price == "vwap":
base_price = pd.Series()
volume = pd.Series()
for inner_indicator in inner_indicators:
base_price = base_price.add(inner_indicator["base_price"] * inner_indicator["volume"], fill_value=0)
volume = volume.add(inner_indicator["volume"], fill_value=0)
base_price /= volume
self["base_price"] = base_price
self["volume"] = volume
else:
raise ValueError(f"base_price {base_price} is not supported!")
self["pa"] = (self["trade_price"] - self["base_price"]) / self["base_price"]
def get_statistics_FFR(self, method="mean"):
if method == "mean":
return self["fulfill_rate"].mean()
elif method == "amount_weighted":
weights = self["deal_amount"].abs()
return (self["fulfill_rate"] * weights).sum() / weights.sum()
elif method == "value_weighted":
weights = (self["deal_amount"] * self["trade_price"]).abs()
return (self["fulfill_rate"] * weights).sum() / weights.sum()
else:
raise ValueError(f"method {method} is not supported!")
def get_statistics_PA(self, method="mean"):
pa_order = self["pa"] * (self["amount"] < 0).astype(int)
if method == "mean":
return pa_order.mean()
elif method == "amount_weighted":
weights = self["deal_amount"].abs()
return (pa_order * weights).sum() / weights.sum()
elif method == "value_weighted":
weights = (self["deal_amount"] * self["trade_price"]).abs()
return (pa_order * weights).sum() / weights.sum()
else:
raise ValueError(f"method {method} is not supported!")
def get_statistics_POS(self):
pa_order = self["pa"] * (self["amount"] < 0).astype(int)
return (pa_order > 1e-8).astype(int).sum() / len(pa_order)

7
qlib/backtest/utils.py
View File

@@ -74,7 +74,12 @@ class TradeCalendarManager:
def get_step_time(self, trade_step=0, shift=0):
"""
Get the time range of trading step
Get the left and right endpoints of the trade_step'th trading interval
About the endpoints:
- Qlib uses the closed interval in time-series data selection, which has the same performance as pandas.Series.loc
- The returned right endpoints should minus 1 seconds becasue of the closed interval representation in Qlib.
Note: Qlib supports up to minutely decision execution, so 1 seconds is less than any trading time interval.
Parameters
----------

393
qlib/contrib/model/pytorch_tcts.py Normal file
View File

@@ -0,0 +1,393 @@
# Copyright (c) Microsoft Corporation.
# Licensed under the MIT License.
from __future__ import division
from __future__ import print_function
import os
import numpy as np
import pandas as pd
import copy
from sklearn.metrics import roc_auc_score, mean_squared_error
import logging
from ...utils import (
unpack_archive_with_buffer,
save_multiple_parts_file,
create_save_path,
drop_nan_by_y_index,
)
from ...log import get_module_logger, TimeInspector
import torch
import torch.nn as nn
import torch.optim as optim
from ...model.base import Model
from ...data.dataset import DatasetH
from ...data.dataset.handler import DataHandlerLP
class TCTS(Model):
"""TCTS Model
Parameters
----------
d_feat : int
input dimension for each time step
metric: str
the evaluate metric used in early stop
optimizer : str
optimizer name
GPU : str
the GPU ID(s) used for training
"""
def __init__(
self,
d_feat=6,
hidden_size=64,
num_layers=2,
dropout=0.0,
n_epochs=200,
batch_size=2000,
early_stop=20,
loss="mse",
fore_optimizer="adam",
weight_optimizer="adam",
output_dim=5,
fore_lr=5e-7,
weight_lr=5e-7,
steps=3,
GPU=0,
seed=None,
target_label=0,
**kwargs
):
# Set logger.
self.logger = get_module_logger("TCTS")
self.logger.info("TCTS pytorch version...")
# set hyper-parameters.
self.d_feat = d_feat
self.hidden_size = hidden_size
self.num_layers = num_layers
self.dropout = dropout
self.n_epochs = n_epochs
self.batch_size = batch_size
self.early_stop = early_stop
self.loss = loss
self.device = torch.device("cuda:%d" % (GPU) if torch.cuda.is_available() else "cpu")
self.use_gpu = torch.cuda.is_available()
self.seed = seed
self.output_dim = output_dim
self.fore_lr = fore_lr
self.weight_lr = weight_lr
self.steps = steps
self.target_label = target_label
self.logger.info(
"TCTS parameters setting:"
"\nd_feat : {}"
"\nhidden_size : {}"
"\nnum_layers : {}"
"\ndropout : {}"
"\nn_epochs : {}"
"\nbatch_size : {}"
"\nearly_stop : {}"
"\nloss_type : {}"
"\nvisible_GPU : {}"
"\nuse_GPU : {}"
"\nseed : {}".format(
d_feat,
hidden_size,
num_layers,
dropout,
n_epochs,
batch_size,
early_stop,
loss,
GPU,
self.use_gpu,
seed,
)
)
if self.seed is not None:
np.random.seed(self.seed)
torch.manual_seed(self.seed)
self.fore_model = GRUModel(
d_feat=self.d_feat,
hidden_size=self.hidden_size,
num_layers=self.num_layers,
dropout=self.dropout,
)
self.weight_model = MLPModel(
d_feat=360 + 2 * self.output_dim + 1,
hidden_size=self.hidden_size,
num_layers=self.num_layers,
dropout=self.dropout,
output_dim=self.output_dim,
)
if fore_optimizer.lower() == "adam":
self.fore_optimizer = optim.Adam(self.fore_model.parameters(), lr=self.fore_lr)
elif fore_optimizer.lower() == "gd":
self.fore_optimizer = optim.SGD(self.fore_model.parameters(), lr=self.fore_lr)
else:
raise NotImplementedError("optimizer {} is not supported!".format(fore_optimizer))
if weight_optimizer.lower() == "adam":
self.weight_optimizer = optim.Adam(self.weight_model.parameters(), lr=self.weight_lr)
elif weight_optimizer.lower() == "gd":
self.weight_optimizer = optim.SGD(self.weight_model.parameters(), lr=self.weight_lr)
else:
raise NotImplementedError("optimizer {} is not supported!".format(weight_optimizer))
self.fitted = False
self.fore_model.to(self.device)
self.weight_model.to(self.device)
def loss_fn(self, pred, label, weight):
loc = torch.argmax(weight, 1)
loss = (pred - label[np.arange(weight.shape[0]), loc]) ** 2
return torch.mean(loss)
def train_epoch(self, x_train, y_train, x_valid, y_valid):
x_train_values = x_train.values
y_train_values = np.squeeze(y_train.values)
indices = np.arange(len(x_train_values))
np.random.shuffle(indices)
init_fore_model = copy.deepcopy(self.fore_model)
for p in init_fore_model.parameters():
p.init_fore_model = False
self.fore_model.train()
self.weight_model.train()
for p in self.weight_model.parameters():
p.requires_grad = False
for p in self.fore_model.parameters():
p.requires_grad = True
for i in range(self.steps):
for i in range(len(indices))[:: self.batch_size]:
if len(indices) - i < self.batch_size:
break
feature = torch.from_numpy(x_train_values[indices[i : i + self.batch_size]]).float().to(self.device)
label = torch.from_numpy(y_train_values[indices[i : i + self.batch_size]]).float().to(self.device)
init_pred = init_fore_model(feature)
pred = self.fore_model(feature)
dis = init_pred - label.transpose(0, 1)
weight_feature = torch.cat((feature, dis.transpose(0, 1), label, init_pred.view(-1, 1)), 1)
weight = self.weight_model(weight_feature)
loss = self.loss_fn(pred, label, weight) # hard
self.fore_optimizer.zero_grad()
loss.backward()
torch.nn.utils.clip_grad_value_(self.fore_model.parameters(), 3.0)
self.fore_optimizer.step()
x_valid_values = x_valid.values
y_valid_values = np.squeeze(y_valid.values)
indices = np.arange(len(x_valid_values))
np.random.shuffle(indices)
for p in self.weight_model.parameters():
p.requires_grad = True
for p in self.fore_model.parameters():
p.requires_grad = False
# fix forecasting model and valid weight model
for i in range(len(indices))[:: self.batch_size]:
if len(indices) - i < self.batch_size:
break
feature = torch.from_numpy(x_valid_values[indices[i : i + self.batch_size]]).float().to(self.device)
label = torch.from_numpy(y_valid_values[indices[i : i + self.batch_size]]).float().to(self.device)
pred = self.fore_model(feature)
dis = pred - label.transpose(0, 1)
weight_feature = torch.cat((feature, dis.transpose(0, 1), label, pred.view(-1, 1)), 1)
weight = self.weight_model(weight_feature)
loc = torch.argmax(weight, 1)
valid_loss = torch.mean((pred - label[:, 0]) ** 2)
loss = torch.mean(-valid_loss * torch.log(weight[np.arange(weight.shape[0]), loc]))
self.weight_optimizer.zero_grad()
loss.backward()
torch.nn.utils.clip_grad_value_(self.weight_model.parameters(), 3.0)
self.weight_optimizer.step()
def test_epoch(self, data_x, data_y):
# prepare training data
x_values = data_x.values
y_values = np.squeeze(data_y.values)
self.fore_model.eval()
scores = []
losses = []
indices = np.arange(len(x_values))
for i in range(len(indices))[:: self.batch_size]:
if len(indices) - i < self.batch_size:
break
feature = torch.from_numpy(x_values[indices[i : i + self.batch_size]]).float().to(self.device)
label = torch.from_numpy(y_values[indices[i : i + self.batch_size]]).float().to(self.device)
pred = self.fore_model(feature)
loss = torch.mean((pred - label[:, abs(self.target_label)]) ** 2)
losses.append(loss.item())
return np.mean(losses)
def fit(
self,
dataset: DatasetH,
evals_result=dict(),
verbose=True,
save_path=None,
):
df_train, df_valid, df_test = dataset.prepare(
["train", "valid", "test"],
col_set=["feature", "label"],
data_key=DataHandlerLP.DK_L,
)
x_train, y_train = df_train["feature"], df_train["label"]
x_valid, y_valid = df_valid["feature"], df_valid["label"]
x_test, y_test = df_test["feature"], df_test["label"]
if save_path == None:
save_path = create_save_path(save_path)
best_loss = np.inf
best_epoch = 0
stop_round = 0
fore_best_param = copy.deepcopy(self.fore_optimizer.state_dict())
weight_best_param = copy.deepcopy(self.weight_optimizer.state_dict())
for epoch in range(self.n_epochs):
print("Epoch:", epoch)
print("training...")
self.train_epoch(x_train, y_train, x_valid, y_valid)
print("evaluating...")
val_loss = self.test_epoch(x_valid, y_valid)
test_loss = self.test_epoch(x_test, y_test)
print("valid %.6f, test %.6f" % (val_loss, test_loss))
if val_loss < best_loss:
best_loss = val_loss
stop_round = 0
best_epoch = epoch
torch.save(copy.deepcopy(self.fore_model.state_dict()), save_path + "_fore_model.bin")
torch.save(copy.deepcopy(self.weight_model.state_dict()), save_path + "_weight_model.bin")
else:
stop_round += 1
if stop_round >= self.early_stop:
print("early stop")
break
print("best loss:", best_loss, "@", best_epoch)
best_param = torch.load(save_path + "_fore_model.bin")
self.fore_model.load_state_dict(best_param)
best_param = torch.load(save_path + "_weight_model.bin")
self.weight_model.load_state_dict(best_param)
self.fitted = True
if self.use_gpu:
torch.cuda.empty_cache()
def predict(self, dataset):
if not self.fitted:
raise ValueError("model is not fitted yet!")
x_test = dataset.prepare("test", col_set="feature")
index = x_test.index
self.fore_model.eval()
x_values = x_test.values
sample_num = x_values.shape[0]
preds = []
for begin in range(sample_num)[:: self.batch_size]:
if sample_num - begin < self.batch_size:
end = sample_num
else:
end = begin + self.batch_size
x_batch = torch.from_numpy(x_values[begin:end]).float().to(self.device)
with torch.no_grad():
if self.use_gpu:
pred = self.fore_model(x_batch).detach().cpu().numpy()
else:
pred = self.fore_model(x_batch).detach().numpy()
preds.append(pred)
return pd.Series(np.concatenate(preds), index=index)
class MLPModel(nn.Module):
def __init__(self, d_feat, hidden_size=256, num_layers=3, dropout=0.0, output_dim=1):
super().__init__()
self.mlp = nn.Sequential()
self.softmax = nn.Softmax(dim=1)
for i in range(num_layers):
if i > 0:
self.mlp.add_module("drop_%d" % i, nn.Dropout(dropout))
self.mlp.add_module("fc_%d" % i, nn.Linear(d_feat if i == 0 else hidden_size, hidden_size))
self.mlp.add_module("relu_%d" % i, nn.ReLU())
self.mlp.add_module("fc_out", nn.Linear(hidden_size, output_dim))
def forward(self, x):
# feature
# [N, F]
out = self.mlp(x).squeeze()
out = self.softmax(out)
return out
class GRUModel(nn.Module):
def __init__(self, d_feat=6, hidden_size=64, num_layers=2, dropout=0.0):
super().__init__()
self.rnn = nn.GRU(
input_size=d_feat,
hidden_size=hidden_size,
num_layers=num_layers,
batch_first=True,
dropout=dropout,
)
self.fc_out = nn.Linear(hidden_size, 1)
self.d_feat = d_feat
def forward(self, x):
# x: [N, F*T]
x = x.reshape(len(x), self.d_feat, -1) # [N, F, T]
x = x.permute(0, 2, 1) # [N, T, F]
out, _ = self.rnn(x)
return self.fc_out(out[:, -1, :]).squeeze()

2
qlib/contrib/model/xgboost.py
View File

@@ -62,7 +62,7 @@ class XGBModel(Model, FeatureInt):
if self.model is None:
raise ValueError("model is not fitted yet!")
x_test = dataset.prepare(segment, col_set="feature", data_key=DataHandlerLP.DK_I)
return pd.Series(self.model.predict(xgb.DMatrix(x_test.values)), index=x_test.index)
return pd.Series(self.model.predict(xgb.DMatrix(x_test)), index=x_test.index)
def get_feature_importance(self, *args, **kwargs) -> pd.Series:
"""get feature importance

26
qlib/contrib/strategy/model_strategy.py
View File

@@ -51,6 +51,11 @@ class TopkDropoutStrategy(ModelStrategy):
trade_exchange : Exchange
exchange that provides market info, used to deal order and generate report
- If `trade_exchange` is None, self.trade_exchange will be set with common_infra
- It allowes different trade_exchanges is used in different executions.
- For example:
- In daily execution, both daily exchange and minutely are usable, but the daily exchange is recommended because it run faster.
- In minutely execution, the daily exchange is not usable, only the minutely exchange is recommended.
"""
super(TopkDropoutStrategy, self).__init__(
model, dataset, level_infra=level_infra, common_infra=common_infra, **kwargs
@@ -253,6 +258,15 @@ class WeightStrategyBase(ModelStrategy):
common_infra=None,
**kwargs,
):
"""
trade_exchange : Exchange
exchange that provides market info, used to deal order and generate report
- If `trade_exchange` is None, self.trade_exchange will be set with common_infra
- It allowes different trade_exchanges is used in different executions.
- For example:
- In daily execution, both daily exchange and minutely are usable, but the daily exchange is recommended because it run faster.
- In minutely execution, the daily exchange is not usable, only the minutely exchange is recommended.
"""
super(WeightStrategyBase, self).__init__(
model, dataset, level_infra=level_infra, common_infra=common_infra, **kwargs
)
@@ -301,18 +315,6 @@ class WeightStrategyBase(ModelStrategy):
raise NotImplementedError()
def generate_trade_decision(self, execute_result=None):
"""
Parameters
-----------
score_series : pd.Seires
stock_id , score.
current : Position()
current of account.
trade_exchange : Exchange()
exchange.
trade_date : pd.Timestamp
date.
"""
# generate_trade_decision
# generate_target_weight_position() and generate_order_list_from_target_weight_position() to generate order_list

216
qlib/contrib/strategy/rule_strategy.py
View File

@@ -1,4 +1,6 @@
import warnings
import numpy as np
import pandas as pd
from typing import List, Union
from ...utils.resam import resam_ts_data
@@ -28,6 +30,10 @@ class TWAPStrategy(BaseStrategy):
trade_exchange : Exchange
exchange that provides market info, used to deal order and generate report
- If `trade_exchange` is None, self.trade_exchange will be set with common_infra
- It allowes different trade_exchanges is used in different executions.
- For example:
- In daily execution, both daily exchange and minutely are usable, but the daily exchange is recommended because it run faster.
- In minutely execution, the daily exchange is not usable, only the minutely exchange is recommended.
"""
super(TWAPStrategy, self).__init__(
@@ -88,27 +94,29 @@ class TWAPStrategy(BaseStrategy):
# considering trade unit
if _amount_trade_unit is None:
# divide the order into equal parts, and trade one part
_order_amount = self.trade_amount[(order.stock_id, order.direction)] / (trade_len - trade_step + 1)
_order_amount = self.trade_amount[(order.stock_id, order.direction)] / (trade_len - trade_step)
# without considering trade unit
elif self.trade_amount[(order.stock_id, order.direction)] >= _amount_trade_unit:
else:
# divide the order into equal parts, and trade one part
# calculate the total count of trade units to trade
trade_unit_cnt = int(self.trade_amount[(order.stock_id, order.direction)] // _amount_trade_unit)
# calculate the amount of one part, ceil the amount
# floor((trade_unit_cnt + trade_len - trade_step) / (trade_len - trade_step + 1)) == ceil(trade_unit_cnt / (trade_len - trade_step + 1))
_order_amount = (
(trade_unit_cnt + trade_len - trade_step) // (trade_len - trade_step + 1) * _amount_trade_unit
(trade_unit_cnt + trade_len - trade_step - 1) // (trade_len - trade_step) * _amount_trade_unit
)
if order.direction == order.SELL:
# sell all amount at last
if self.trade_amount[(order.stock_id, order.direction)] > 1e-5 and (
_order_amount is None or trade_step == trade_len
_order_amount < 1e-5 or trade_step == trade_len - 1
):
_order_amount = self.trade_amount[(order.stock_id, order.direction)]
if _order_amount:
_order_amount = min(_order_amount, self.trade_amount[(order.stock_id, order.direction)])
_order_amount = min(_order_amount, self.trade_amount[(order.stock_id, order.direction)])
if _order_amount > 1e-5:
_order = Order(
stock_id=order.stock_id,
amount=_order_amount,
@@ -145,6 +153,10 @@ class SBBStrategyBase(BaseStrategy):
trade_exchange : Exchange
exchange that provides market info, used to deal order and generate report
- If `trade_exchange` is None, self.trade_exchange will be set with common_infra
- It allowes different trade_exchanges is used in different executions.
- For example:
- In daily execution, both daily exchange and minutely are usable, but the daily exchange is recommended because it run faster.
- In minutely execution, the daily exchange is not usable, only the minutely exchange is recommended.
"""
super(SBBStrategyBase, self).__init__(
outer_trade_decision=outer_trade_decision, level_infra=level_infra, common_infra=common_infra
@@ -222,7 +234,7 @@ class SBBStrategyBase(BaseStrategy):
# divide the order into equal parts, and trade one part
_order_amount = self.trade_amount[(order.stock_id, order.direction)] / (trade_len - trade_step)
# without considering trade unit
elif self.trade_amount[(order.stock_id, order.direction)] >= _amount_trade_unit:
else:
# divide the order into equal parts, and trade one part
# calculate the total count of trade units to trade
trade_unit_cnt = int(self.trade_amount[(order.stock_id, order.direction)] // _amount_trade_unit)
@@ -234,11 +246,13 @@ class SBBStrategyBase(BaseStrategy):
if order.direction == order.SELL:
# sell all amount at last
if self.trade_amount[(order.stock_id, order.direction)] > 1e-5 and (
_order_amount is None or trade_step == trade_len - 1
_order_amount < 1e-5 or trade_step == trade_len - 1
):
_order_amount = self.trade_amount[(order.stock_id, order.direction)]
if _order_amount:
_order_amount = min(_order_amount, self.trade_amount[(order.stock_id, order.direction)])
if _order_amount > 1e-5:
_order = Order(
stock_id=order.stock_id,
amount=_order_amount,
@@ -258,7 +272,7 @@ class SBBStrategyBase(BaseStrategy):
2 * self.trade_amount[(order.stock_id, order.direction)] / (trade_len - trade_step + 1)
)
# without considering trade unit
elif self.trade_amount[(order.stock_id, order.direction)] >= _amount_trade_unit:
else:
# cal how many trade unit
trade_unit_cnt = int(self.trade_amount[(order.stock_id, order.direction)] // _amount_trade_unit)
# N trade day left, divide the order into N + 1 parts, and trade 2 parts
@@ -270,13 +284,14 @@ class SBBStrategyBase(BaseStrategy):
)
if order.direction == order.SELL:
# sell all amount at last
if self.trade_amount[(order.stock_id, order.direction)] >= 1e-5 and (
_order_amount is None or trade_step == trade_len - 1
if self.trade_amount[(order.stock_id, order.direction)] > 1e-5 and (
_order_amount < 1e-5 or trade_step == trade_len - 1
):
_order_amount = self.trade_amount[(order.stock_id, order.direction)]
if _order_amount:
_order_amount = min(_order_amount, self.trade_amount[(order.stock_id, order.direction)])
_order_amount = min(_order_amount, self.trade_amount[(order.stock_id, order.direction)])
if _order_amount > 1e-5:
if trade_step % 2 == 0:
# in the first one of two adjacent bars
# if look short on the price, sell the stock more
@@ -402,3 +417,176 @@ class SBBStrategyEMA(SBBStrategyBase):
# if EMA signal > 0, return short trend
else:
return self.TREND_SHORT
class VAStrategy(BaseStrategy):
def __init__(
self,
lamb: float = 1e-6,
eta: float = 2.5e-6,
window_size: int = 20,
outer_trade_decision: List[Order] = None,
instruments: Union[List, str] = "csi300",
freq: str = "day",
trade_exchange: Exchange = None,
level_infra: LevelInfrastructure = None,
common_infra: CommonInfrastructure = None,
**kwargs,
):
"""
Parameters
----------
instruments : Union[List, str], optional
instruments of Volatility, by default "csi300"
freq : str, optional
freq of Volatility, by default "day"
Note: `freq` may be different from `time_per_step`
"""
self.lamb = lamb
self.eta = eta
self.window_size = window_size
if instruments is None:
warnings.warn("`instruments` is not set, will load all stocks")
self.instruments = "all"
if isinstance(instruments, str):
self.instruments = D.instruments(instruments)
self.freq = freq
super(VAStrategy, self).__init__(outer_trade_decision, level_infra, common_infra, **kwargs)
if trade_exchange is not None:
self.trade_exchange = trade_exchange
def _reset_signal(self):
trade_len = self.trade_calendar.get_trade_len()
fields = [
f"Power(Sum(Power(Log($close/Ref($close, 1)), 2), {self.window_size})/{self.window_size - 1}-Power(Sum(Log($close/Ref($close, 1)), {self.window_size}), 2)/({self.window_size}*{self.window_size - 1}), 0.5)"
]
signal_start_time, _ = self.trade_calendar.get_step_time(trade_step=0, shift=1)
_, signal_end_time = self.trade_calendar.get_step_time(trade_step=trade_len - 1, shift=1)
signal_df = D.features(
self.instruments, fields, start_time=signal_start_time, end_time=signal_end_time, freq=self.freq
)
signal_df = convert_index_format(signal_df)
signal_df.columns = ["volatility"]
self.signal = {}
if not signal_df.empty:
for stock_id, stock_val in signal_df.groupby(level="instrument"):
self.signal[stock_id] = stock_val
def reset_common_infra(self, common_infra):
"""
Parameters
----------
common_infra : CommonInfrastructure, optional
common infrastructure for backtesting, by default None
- It should include `trade_account`, used to get position
- It should include `trade_exchange`, used to provide market info
"""
super(VAStrategy, self).reset_common_infra(common_infra)
if common_infra.has("trade_exchange"):
self.trade_exchange = common_infra.get("trade_exchange")
def reset_level_infra(self, level_infra):
"""
reset level-shared infra
- After reset the trade calendar, the signal will be changed
"""
if not hasattr(self, "level_infra"):
self.level_infra = level_infra
else:
self.level_infra.update(level_infra)
if level_infra.has("trade_calendar"):
self.trade_calendar = level_infra.get("trade_calendar")
self._reset_signal()
def reset(self, outer_trade_decision: List[Order] = None, **kwargs):
"""
Parameters
----------
outer_trade_decision : List[Order], optional
"""
super(VAStrategy, self).reset(outer_trade_decision=outer_trade_decision, **kwargs)
if outer_trade_decision is not None:
self.trade_amount = {}
# init the trade amount of order and predicted trade trend
for order in outer_trade_decision:
self.trade_amount[(order.stock_id, order.direction)] = order.amount
def generate_trade_decision(self, execute_result=None):
# update the order amount
if execute_result is not None:
for order, _, _, _ in execute_result:
self.trade_amount[(order.stock_id, order.direction)] -= order.deal_amount
# get the number of trading step finished, trade_step can be [0, 1, 2, ..., trade_len - 1]
trade_step = self.trade_calendar.get_trade_step()
# get the total count of trading step
trade_len = self.trade_calendar.get_trade_len()
trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
order_list = []
for order in self.outer_trade_decision:
# if not tradable, continue
if not self.trade_exchange.is_stock_tradable(
stock_id=order.stock_id, start_time=trade_start_time, end_time=trade_end_time
):
continue
_order_amount = None
# considering trade unit
sig_sam = (
resam_ts_data(self.signal[order.stock_id]["volatility"], pred_start_time, pred_end_time, method="last")
if order.stock_id in self.signal
else None
)
if sig_sam is None or sig_sam.iloc[0] is None:
# no signal, TWAP
_amount_trade_unit = self.trade_exchange.get_amount_of_trade_unit(order.factor)
if _amount_trade_unit is None:
# divide the order into equal parts, and trade one part
_order_amount = self.trade_amount[(order.stock_id, order.direction)] / (trade_len - trade_step)
else:
# divide the order into equal parts, and trade one part
# calculate the total count of trade units to trade
trade_unit_cnt = int(self.trade_amount[(order.stock_id, order.direction)] // _amount_trade_unit)
# calculate the amount of one part, ceil the amount
# floor((trade_unit_cnt + trade_len - trade_step - 1) / (trade_len - trade_step)) == ceil(trade_unit_cnt / (trade_len - trade_step))
_order_amount = (
(trade_unit_cnt + trade_len - trade_step - 1) // (trade_len - trade_step) * _amount_trade_unit
)
else:
# VA strategy
kappa_tild = self.lamb / self.eta * sig_sam.iloc[0] * sig_sam.iloc[0]
kappa = np.arccosh(kappa_tild / 2 + 1)
amount_ratio = (
np.sinh(kappa * (trade_len - trade_step)) - np.sinh(kappa * (trade_len - trade_step - 1))
) / np.sinh(kappa * trade_len)
_order_amount = order.amount * amount_ratio
_order_amount = self.trade_exchange.round_amount_by_trade_unit(_order_amount, order.factor)
if order.direction == order.SELL:
# sell all amount at last
if self.trade_amount[(order.stock_id, order.direction)] > 1e-5 and (
_order_amount < 1e-5 or trade_step == trade_len - 1
):
_order_amount = self.trade_amount[(order.stock_id, order.direction)]
_order_amount = min(_order_amount, self.trade_amount[(order.stock_id, order.direction)])
if _order_amount > 1e-5:
_order = Order(
stock_id=order.stock_id,
amount=_order_amount,
start_time=trade_start_time,
end_time=trade_end_time,
direction=order.direction, # 1 for buy
factor=order.factor,
)
order_list.append(_order)
return order_list

86
qlib/data/data.py
View File

@@ -65,7 +65,6 @@ class CalendarProvider(abc.ABC, ProviderBackendMixin):
def __init__(self, *args, **kwargs):
self.backend = kwargs.get("backend", {})
@abc.abstractmethod
def calendar(self, start_time=None, end_time=None, freq="day", freq_sam=None, future=False):
"""Get calendar of certain market in given time range.
@@ -87,7 +86,22 @@ class CalendarProvider(abc.ABC, ProviderBackendMixin):
list
calendar list
"""
raise NotImplementedError("Subclass of CalendarProvider must implement `calendar` method")
_calendar, _ = self._get_calendar(freq=freq, freq_sam=freq_sam, future=future)
# strip
if start_time:
start_time = pd.Timestamp(start_time)
if start_time > _calendar[-1]:
return np.array([])
else:
start_time = _calendar[0]
if end_time:
end_time = pd.Timestamp(end_time)
if end_time < _calendar[0]:
return np.array([])
else:
end_time = _calendar[-1]
st, et, si, ei = self.locate_index(start_time, end_time, freq=freq, freq_sam=freq_sam, future=future)
return _calendar[si : ei + 1]
def locate_index(self, start_time, end_time, freq, freq_sam=None, future=False):
"""Locate the start time index and end time index in a calendar under certain frequency.
@@ -172,6 +186,21 @@ class CalendarProvider(abc.ABC, ProviderBackendMixin):
"""Get the uri of calendar generation task."""
return hash_args(start_time, end_time, freq, future)
def load_calendar(self, freq, future):
"""Load original calendar timestamp from file.
Parameters
----------
freq : str
frequency of read calendar file.
Returns
----------
list
list of timestamps
"""
raise NotImplementedError("Subclass of CalendarProvider must implement `load_calendar` method")
class InstrumentProvider(abc.ABC, ProviderBackendMixin):
"""Instrument provider base class
@@ -457,7 +486,8 @@ class DatasetProvider(abc.ABC):
normalize_column_names = normalize_cache_fields(column_names)
data = dict()
# One process for one task, so that the memory will be freed quicker.
workers = min(C.kernels, len(instruments_d))
workers = max(min(C.kernels, len(instruments_d)), 1)
if C.maxtasksperchild is None:
p = Pool(processes=workers)
else:
@@ -504,7 +534,9 @@ class DatasetProvider(abc.ABC):
data = pd.concat(new_data, names=["instrument"], sort=False)
data = DiskDatasetCache.cache_to_origin_data(data, column_names)
else:
data = pd.DataFrame(columns=column_names)
data = pd.DataFrame(
index=pd.MultiIndex.from_arrays([[], []], names=("instrument", "datetime")), columns=column_names
)
return data
@@ -558,19 +590,6 @@ class LocalCalendarProvider(CalendarProvider):
return os.path.join(C.get_data_path(), "calendars", "{}.txt")
def load_calendar(self, freq, future):
"""Load original calendar timestamp from file.
Parameters
----------
freq : str
frequency of read calendar file.
Returns
----------
list
list of timestamps
"""
try:
backend_obj = self.backend_obj(freq=freq, future=future).data
except ValueError:
@@ -587,24 +606,6 @@ class LocalCalendarProvider(CalendarProvider):
return [pd.Timestamp(x) for x in backend_obj]
def calendar(self, start_time=None, end_time=None, freq="day", freq_sam=None, future=False):
_calendar, _ = self._get_calendar(freq=freq, freq_sam=freq_sam, future=future)
# strip
if start_time:
start_time = pd.Timestamp(start_time)
if start_time > _calendar[-1]:
return np.array([])
else:
start_time = _calendar[0]
if end_time:
end_time = pd.Timestamp(end_time)
if end_time < _calendar[0]:
return np.array([])
else:
end_time = _calendar[-1]
st, et, si, ei = self.locate_index(start_time, end_time, freq=freq, freq_sam=freq_sam, future=future)
return _calendar[si : ei + 1]
class LocalInstrumentProvider(InstrumentProvider):
"""Local instrument data provider class
@@ -719,7 +720,9 @@ class LocalDatasetProvider(DatasetProvider):
column_names = self.get_column_names(fields)
cal = Cal.calendar(start_time, end_time, freq)
if len(cal) == 0:
return pd.DataFrame(columns=column_names)
return pd.DataFrame(
index=pd.MultiIndex.from_arrays([[], []], names=("instrument", "datetime")), columns=column_names
)
start_time = cal[0]
end_time = cal[-1]
@@ -741,7 +744,7 @@ class LocalDatasetProvider(DatasetProvider):
return
start_time = cal[0]
end_time = cal[-1]
workers = min(C.kernels, len(instruments_d))
workers = max(min(C.kernels, len(instruments_d)), 1)
if C.maxtasksperchild is None:
p = Pool(processes=workers)
else:
@@ -789,7 +792,7 @@ class ClientCalendarProvider(CalendarProvider):
def calendar(self, start_time=None, end_time=None, freq="day", freq_sam=None, future=False):
self.conn.send_request(
request_type="trade_calendar",
request_type="calendar",
request_content={
"start_time": str(start_time),
"end_time": str(end_time),
@@ -902,7 +905,10 @@ class ClientDatasetProvider(DatasetProvider):
column_names = self.get_column_names(fields)
cal = Cal.calendar(start_time, end_time, freq)
if len(cal) == 0:
return pd.DataFrame(columns=column_names)
return pd.DataFrame(
index=pd.MultiIndex.from_arrays([[], []], names=("instrument", "datetime")),
columns=column_names,
)
start_time = cal[0]
end_time = cal[-1]
@@ -1004,7 +1010,7 @@ class LocalProvider(BaseProvider):
:param type: The type of resource for the uri
:param **kwargs:
"""
if type == "trade_calendar":
if type == "calendar":
return Cal._uri(**kwargs)
elif type == "instrument":
return Inst._uri(**kwargs)

81
qlib/model/trainer.py
View File

@@ -12,9 +12,11 @@ In ``DelayTrainer``, the first step is only to save some necessary info to model
"""
import socket
import time
from typing import Callable, List
from qlib.data.dataset import Dataset
from qlib.log import get_module_logger
from qlib.model.base import Model
from qlib.utils import flatten_dict, get_cls_kwargs, init_instance_by_config
from qlib.workflow import R
@@ -190,6 +192,8 @@ class TrainerR(Trainer):
Returns:
List[Recorder]: a list of Recorders
"""
if isinstance(tasks, dict):
tasks = [tasks]
if len(tasks) == 0:
return []
if train_func is None:
@@ -213,6 +217,8 @@ class TrainerR(Trainer):
Returns:
List[Recorder]: the same list as the param.
"""
if isinstance(recs, Recorder):
recs = [recs]
for rec in recs:
rec.set_tags(**{self.STATUS_KEY: self.STATUS_END})
return recs
@@ -250,6 +256,8 @@ class DelayTrainerR(TrainerR):
Returns:
List[Recorder]: a list of Recorders
"""
if isinstance(recs, Recorder):
recs = [recs]
if end_train_func is None:
end_train_func = self.end_train_func
if experiment_name is None:
@@ -275,6 +283,9 @@ class TrainerRM(Trainer):
STATUS_BEGIN = "begin_task_train"
STATUS_END = "end_task_train"
# This tag is the _id in TaskManager to distinguish tasks.
TM_ID = "_id in TaskManager"
def __init__(self, experiment_name: str = None, task_pool: str = None, train_func=task_train):
"""
Init TrainerR.
@@ -315,6 +326,8 @@ class TrainerRM(Trainer):
Returns:
List[Recorder]: a list of Recorders
"""
if isinstance(tasks, dict):
tasks = [tasks]
if len(tasks) == 0:
return []
if train_func is None:
@@ -326,19 +339,25 @@ class TrainerRM(Trainer):
task_pool = experiment_name
tm = TaskManager(task_pool=task_pool)
_id_list = tm.create_task(tasks) # all tasks will be saved to MongoDB
query = {"_id": {"$in": _id_list}}
run_task(
train_func,
task_pool,
query=query, # only train these tasks
experiment_name=experiment_name,
before_status=before_status,
after_status=after_status,
**kwargs,
)
if not self.is_delay():
tm.wait(query=query)
recs = []
for _id in _id_list:
rec = tm.re_query(_id)["res"]
rec.set_tags(**{self.STATUS_KEY: self.STATUS_BEGIN})
rec.set_tags(**{self.TM_ID: _id})
recs.append(rec)
return recs
@@ -352,10 +371,33 @@ class TrainerRM(Trainer):
Returns:
List[Recorder]: the same list as the param.
"""
if isinstance(recs, Recorder):
recs = [recs]
for rec in recs:
rec.set_tags(**{self.STATUS_KEY: self.STATUS_END})
return recs
def worker(
self,
train_func: Callable = None,
experiment_name: str = None,
):
"""
The multiprocessing method for `train`. It can share a same task_pool with `train` and can run in other progress or other machines.
Args:
train_func (Callable): the training method which needs at least `task`s and `experiment_name`. None for the default training method.
experiment_name (str): the experiment name, None for use default name.
"""
if train_func is None:
train_func = self.train_func
if experiment_name is None:
experiment_name = self.experiment_name
task_pool = self.task_pool
if task_pool is None:
task_pool = experiment_name
run_task(train_func, task_pool=task_pool, experiment_name=experiment_name)
class DelayTrainerRM(TrainerRM):
"""
@@ -395,6 +437,8 @@ class DelayTrainerRM(TrainerRM):
Returns:
List[Recorder]: a list of Recorders
"""
if isinstance(tasks, dict):
tasks = [tasks]
if len(tasks) == 0:
return []
return super().train(
@@ -410,8 +454,6 @@ class DelayTrainerRM(TrainerRM):
Given a list of Recorder and return a list of trained Recorder.
This class will finish real data loading and model fitting.
NOTE: This method will train all STATUS_PART_DONE tasks in the task pool, not only the ``recs``.
Args:
recs (list): a list of Recorder, the tasks have been saved to them.
end_train_func (Callable, optional): the end_train method which need at least `recorder`s and `experiment_name`. Defaults to None for using self.end_train_func.
@@ -421,7 +463,8 @@ class DelayTrainerRM(TrainerRM):
Returns:
List[Recorder]: a list of Recorders
"""
if isinstance(recs, Recorder):
recs = [recs]
if end_train_func is None:
end_train_func = self.end_train_func
if experiment_name is None:
@@ -429,18 +472,44 @@ class DelayTrainerRM(TrainerRM):
task_pool = self.task_pool
if task_pool is None:
task_pool = experiment_name
tasks = []
_id_list = []
for rec in recs:
tasks.append(rec.load_object("task"))
_id_list.append(rec.list_tags()[self.TM_ID])
query = {"_id": {"$in": _id_list}}
run_task(
end_train_func,
task_pool,
query={"filter": {"$in": tasks}}, # only train these tasks
query=query, # only train these tasks
experiment_name=experiment_name,
before_status=TaskManager.STATUS_PART_DONE,
**kwargs,
)
TaskManager(task_pool=task_pool).wait(query=query)
for rec in recs:
rec.set_tags(**{self.STATUS_KEY: self.STATUS_END})
return recs
def worker(self, end_train_func=None, experiment_name: str = None):
"""
The multiprocessing method for `end_train`. It can share a same task_pool with `end_train` and can run in other progress or other machines.
Args:
end_train_func (Callable, optional): the end_train method which need at least `recorder`s and `experiment_name`. Defaults to None for using self.end_train_func.
experiment_name (str): the experiment name, None for use default name.
"""
if end_train_func is None:
end_train_func = self.end_train_func
if experiment_name is None:
experiment_name = self.experiment_name
task_pool = self.task_pool
if task_pool is None:
task_pool = experiment_name
run_task(
end_train_func,
task_pool=task_pool,
experiment_name=experiment_name,
before_status=TaskManager.STATUS_PART_DONE,
)

6
qlib/rl/interpreter.py
View File

@@ -5,14 +5,14 @@
class BaseInterpreter:
"""Base Interpreter"""
def interpret(**kwargs):
def interpret(self, **kwargs):
raise NotImplementedError("interpret is not implemented!")
class ActionInterpreter(BaseInterpreter):
"""Action Interpreter that interpret rl agent action into qlib orders"""
def interpret(action, **kwargs):
def interpret(self, action, **kwargs):
"""interpret method
Parameters
@@ -32,7 +32,7 @@ class ActionInterpreter(BaseInterpreter):
class StateInterpreter(BaseInterpreter):
"""State Interpreter that interpret execution result of qlib executor into rl env state"""
def interpret(execute_result, **kwargs):
def interpret(self, execute_result, **kwargs):
"""interpret method
Parameters

2
qlib/strategy/base.py
View File

@@ -175,7 +175,7 @@ class RLIntStrategy(RLStrategy):
self.action_interpreter = init_instance_by_config(action_interpreter, accept_types=ActionInterpreter)
def generate_trade_decision(self, execute_result=None):
_interpret_state = self.state_interpretor.interpret(execute_result=execute_result)
_interpret_state = self.state_interpreter.interpret(execute_result=execute_result)
_action = self.policy.step(_interpret_state)
_trade_decision = self.action_interpreter.interpret(action=_action)
return _trade_decision

94
qlib/tests/config.py
View File

@@ -43,17 +43,29 @@ RECORD_CONFIG = [
]
def get_data_handler_config(market=CSI300_MARKET):
def get_data_handler_config(
start_time="2008-01-01",
end_time="2020-08-01",
fit_start_time="2008-01-01",
fit_end_time="2014-12-31",
instruments=CSI300_MARKET,
):
return {
"start_time": "2008-01-01",
"end_time": "2020-08-01",
"fit_start_time": "2008-01-01",
"fit_end_time": "2014-12-31",
"instruments": market,
"start_time": start_time,
"end_time": end_time,
"fit_start_time": fit_start_time,
"fit_end_time": fit_end_time,
"instruments": instruments,
}
def get_dataset_config(market=CSI300_MARKET, dataset_class=DATASET_ALPHA158_CLASS):
def get_dataset_config(
dataset_class=DATASET_ALPHA158_CLASS,
train=("2008-01-01", "2014-12-31"),
valid=("2015-01-01", "2016-12-31"),
test=("2017-01-01", "2020-08-01"),
handler_kwargs={"instruments": CSI300_MARKET},
):
return {
"class": "DatasetH",
"module_path": "qlib.data.dataset",
@@ -61,48 +73,88 @@ def get_dataset_config(market=CSI300_MARKET, dataset_class=DATASET_ALPHA158_CLAS
"handler": {
"class": dataset_class,
"module_path": "qlib.contrib.data.handler",
"kwargs": get_data_handler_config(market),
"kwargs": get_data_handler_config(**handler_kwargs),
},
"segments": {
"train": ("2008-01-01", "2014-12-31"),
"valid": ("2015-01-01", "2016-12-31"),
"test": ("2017-01-01", "2020-08-01"),
"train": train,
"valid": valid,
"test": test,
},
},
}
def get_gbdt_task(market=CSI300_MARKET):
def get_gbdt_task(dataset_kwargs={}, handler_kwargs={"instruments": CSI300_MARKET}):
return {
"model": GBDT_MODEL,
"dataset": get_dataset_config(market),
"dataset": get_dataset_config(**dataset_kwargs, handler_kwargs=handler_kwargs),
}
def get_record_lgb_config(market=CSI300_MARKET):
def get_record_lgb_config(dataset_kwargs={}, handler_kwargs={"instruments": CSI300_MARKET}):
return {
"model": {
"class": "LGBModel",
"module_path": "qlib.contrib.model.gbdt",
},
"dataset": get_dataset_config(market),
"dataset": get_dataset_config(**dataset_kwargs, handler_kwargs=handler_kwargs),
"record": RECORD_CONFIG,
}
def get_record_xgboost_config(market=CSI300_MARKET):
def get_record_xgboost_config(dataset_kwargs={}, handler_kwargs={"instruments": CSI300_MARKET}):
return {
"model": {
"class": "XGBModel",
"module_path": "qlib.contrib.model.xgboost",
},
"dataset": get_dataset_config(market),
"dataset": get_dataset_config(**dataset_kwargs, handler_kwargs=handler_kwargs),
"record": RECORD_CONFIG,
}
CSI300_DATASET_CONFIG = get_dataset_config(market=CSI300_MARKET)
CSI300_GBDT_TASK = get_gbdt_task(market=CSI300_MARKET)
CSI300_DATASET_CONFIG = get_dataset_config(handler_kwargs={"instruments": CSI300_MARKET})
CSI300_GBDT_TASK = get_gbdt_task(handler_kwargs={"instruments": CSI300_MARKET})
CSI100_RECORD_XGBOOST_TASK_CONFIG = get_record_xgboost_config(market=CSI100_MARKET)
CSI100_RECORD_LGB_TASK_CONFIG = get_record_lgb_config(market=CSI100_MARKET)
CSI100_RECORD_XGBOOST_TASK_CONFIG = get_record_xgboost_config(handler_kwargs={"instruments": CSI100_MARKET})
CSI100_RECORD_LGB_TASK_CONFIG = get_record_lgb_config(handler_kwargs={"instruments": CSI100_MARKET})
# use for rolling_online_managment.py
ROLLING_HANDLER_CONFIG = {
"start_time": "2013-01-01",
"end_time": "2020-09-25",
"fit_start_time": "2013-01-01",
"fit_end_time": "2014-12-31",
"instruments": CSI100_MARKET,
}
ROLLING_DATASET_CONFIG = {
"train": ("2013-01-01", "2014-12-31"),
"valid": ("2015-01-01", "2015-12-31"),
"test": ("2016-01-01", "2020-07-10"),
}
CSI100_RECORD_XGBOOST_TASK_CONFIG_ROLLING = get_record_xgboost_config(
dataset_kwargs=ROLLING_DATASET_CONFIG, handler_kwargs=ROLLING_HANDLER_CONFIG
)
CSI100_RECORD_LGB_TASK_CONFIG_ROLLING = get_record_lgb_config(
dataset_kwargs=ROLLING_DATASET_CONFIG, handler_kwargs=ROLLING_HANDLER_CONFIG
)
# use for online_management_simulate.py
ONLINE_HANDLER_CONFIG = {
"start_time": "2018-01-01",
"end_time": "2018-10-31",
"fit_start_time": "2018-01-01",
"fit_end_time": "2018-03-31",
"instruments": CSI100_MARKET,
}
ONLINE_DATASET_CONFIG = {
"train": ("2018-01-01", "2018-03-31"),
"valid": ("2018-04-01", "2018-05-31"),
"test": ("2018-06-01", "2018-09-10"),
}
CSI100_RECORD_XGBOOST_TASK_CONFIG_ONLINE = get_record_xgboost_config(
dataset_kwargs=ONLINE_DATASET_CONFIG, handler_kwargs=ONLINE_HANDLER_CONFIG
)
CSI100_RECORD_LGB_TASK_CONFIG_ONLINE = get_record_lgb_config(
dataset_kwargs=ONLINE_DATASET_CONFIG, handler_kwargs=ONLINE_HANDLER_CONFIG
)

91
qlib/utils/resam.py
View File

@@ -8,6 +8,11 @@ from typing import Tuple, List, Union, Optional, Callable
from . import lazy_sort_index
from ..config import C
NORM_FREQ_MONTH = "month"
NORM_FREQ_WEEK = "week"
NORM_FREQ_DAY = "day"
NORM_FREQ_MINUTE = "minute"
def parse_freq(freq: str) -> Tuple[int, str]:
"""
@@ -43,14 +48,14 @@ def parse_freq(freq: str) -> Tuple[int, str]:
_count = int(match_obj.group(1)) if match_obj.group(1) else 1
_freq = match_obj.group(2)
_freq_format_dict = {
"month": "month",
"mon": "month",
"week": "week",
"w": "week",
"day": "day",
"d": "day",
"minute": "minute",
"min": "minute",
"month": NORM_FREQ_MONTH,
"mon": NORM_FREQ_MONTH,
"week": NORM_FREQ_WEEK,
"w": NORM_FREQ_WEEK,
"day": NORM_FREQ_DAY,
"d": NORM_FREQ_DAY,
"minute": NORM_FREQ_MINUTE,
"min": NORM_FREQ_MINUTE,
}
return _count, _freq_format_dict[_freq]
@@ -81,7 +86,7 @@ def resam_calendar(calendar_raw: np.ndarray, freq_raw: str, freq_sam: str) -> np
return calendar_raw
# if freq_sam is xminute, divide each trading day into several bars evenly
if freq_sam == "minute":
if freq_sam == NORM_FREQ_MINUTE:
def cal_sam_minute(x, sam_minutes):
"""
@@ -114,7 +119,7 @@ def resam_calendar(calendar_raw: np.ndarray, freq_raw: str, freq_sam: str) -> np
else:
raise ValueError("calendar minute_index error, check `min_data_shift` in qlib.config.C")
if freq_raw != "minute":
if freq_raw != NORM_FREQ_MINUTE:
raise ValueError("when sampling minute calendar, freq of raw calendar must be minute or min")
else:
if raw_count > sam_count:
@@ -125,15 +130,15 @@ def resam_calendar(calendar_raw: np.ndarray, freq_raw: str, freq_sam: str) -> np
# else, convert the raw calendar into day calendar, and divide the whole calendar into several bars evenly
else:
_calendar_day = np.unique(list(map(lambda x: pd.Timestamp(x.year, x.month, x.day, 0, 0, 0), calendar_raw)))
if freq_sam == "day":
if freq_sam == NORM_FREQ_DAY:
return _calendar_day[::sam_count]
elif freq_sam == "week":
elif freq_sam == NORM_FREQ_WEEK:
_day_in_week = np.array(list(map(lambda x: x.dayofweek, _calendar_day)))
_calendar_week = _calendar_day[np.ediff1d(_day_in_week, to_begin=-1) < 0]
return _calendar_week[::sam_count]
elif freq_sam == "month":
elif freq_sam == NORM_FREQ_MONTH:
_day_in_month = np.array(list(map(lambda x: x.day, _calendar_day)))
_calendar_month = _calendar_day[np.ediff1d(_day_in_month, to_begin=-1) < 0]
return _calendar_month[::sam_count]
@@ -184,7 +189,7 @@ def get_resam_calendar(
freq, freq_sam = freq, None
except (ValueError, KeyError):
freq_sam = freq
if norm_freq in ["month", "week", "day"]:
if norm_freq in [NORM_FREQ_MONTH, NORM_FREQ_WEEK, NORM_FREQ_DAY]:
try:
_calendar = Cal.calendar(
start_time=start_time, end_time=end_time, freq="day", freq_sam=freq, future=future
@@ -195,7 +200,7 @@ def get_resam_calendar(
start_time=start_time, end_time=end_time, freq="1min", freq_sam=freq, future=future
)
freq = "1min"
elif norm_freq == "minute":
elif norm_freq == NORM_FREQ_MINUTE:
_calendar = Cal.calendar(
start_time=start_time, end_time=end_time, freq="1min", freq_sam=freq, future=future
)
@@ -205,6 +210,57 @@ def get_resam_calendar(
return _calendar, freq, freq_sam
def get_higher_freq_feature(instruments, fields, start_time=None, end_time=None, freq="day", disk_cache=1):
"""[summary]
Parameters
----------
instruments : [type]
[description]
fields : [type]
[description]
start_time : [type], optional
[description], by default None
end_time : [type], optional
[description], by default None
freq : str, optional
[description], by default "day"
disk_cache : int, optional
[description], by default 1
Returns
-------
[type]
[description]
Raises
------
ValueError
[description]
"""
from ..data.data import D
try:
_result = D.features(instruments, fields, start_time, end_time, freq=freq, disk_cache=disk_cache)
_freq = freq
except (ValueError, KeyError):
_, norm_freq = parse_freq(freq)
if norm_freq in [NORM_FREQ_MONTH, NORM_FREQ_WEEK, NORM_FREQ_DAY]:
try:
_result = D.features(instruments, fields, start_time, end_time, freq="day", disk_cache=disk_cache)
_freq = "day"
except (ValueError, KeyError):
_result = D.features(instruments, fields, start_time, end_time, freq="1min", disk_cache=disk_cache)
_freq = "1min"
elif norm_freq == NORM_FREQ_MINUTE:
_result = D.features(instruments, fields, start_time, end_time, freq="1min", disk_cache=disk_cache)
_freq = "1min"
else:
raise ValueError(f"freq {freq} is not supported")
return _result, _freq
def resam_ts_data(
ts_feature: Union[pd.DataFrame, pd.Series],
start_time: Union[str, pd.Timestamp] = None,
@@ -273,8 +329,9 @@ def resam_ts_data(
end sampling time, by default None
method : Union[str, Callable], optional
sample method, apply method function to each stock series data, by default "last"
- If type(method) is str, it should be an attribute of SeriesGroupBy or DataFrameGroupby, and run feature.groupby
- If `feature` has MultiIndex[instrument, datetime], method must be a member of pandas.groupby when it's type is str.or callable function.
- If type(method) is str or callable function, it should be an attribute of SeriesGroupBy or DataFrameGroupby, and applies groupy.method for the sliced time-series data
- If method is None, do nothing for the sliced time-series data.
- Only when the index `feature` is MultiIndex[instrument, datetime], the method is valid.
method_kwargs : dict, optional
arguments of method, by default {}

36
qlib/workflow/online/manager.py
View File

@@ -18,10 +18,12 @@ There are 4 total situations for using different trainers in different situation
========================= ===================================================================================
Situations Description
========================= ===================================================================================
Online + Trainer When you REAL want to do a routine, the Trainer will help you train the models.
Online + Trainer When you want to do a REAL routine, the Trainer will help you train the models. It
will train models task by task and strategy by strategy.
Online + DelayTrainer In normal online routine, whether Trainer or DelayTrainer will REAL train models
in this routine. So it is not necessary to use DelayTrainer when do a REAL routine.
Online + DelayTrainer When your models don't have any temporal dependence, the DelayTrainer will train
nothing until all tasks have been prepared. It makes user can train all tasks in
the end of `routine` or `first_train`.
Simulation + Trainer When your models have some temporal dependence on the previous models, then you
need to consider using Trainer. This means it will REAL train your models in
@@ -103,17 +105,21 @@ class OnlineManager(Serializable):
"""
if strategies is None:
strategies = self.strategies
for strategy in strategies:
models_list = []
for strategy in strategies:
self.logger.info(f"Strategy `{strategy.name_id}` begins first training...")
tasks = strategy.first_tasks()
models = self.trainer.train(tasks, experiment_name=strategy.name_id)
models = self.trainer.end_train(models, experiment_name=strategy.name_id)
models_list.append(models)
self.logger.info(f"Finished training {len(models)} models.")
online_models = strategy.prepare_online_models(models, **model_kwargs)
self.history.setdefault(self.cur_time, {})[strategy] = online_models
if not self.status == self.STATUS_SIMULATING or not self.trainer.is_delay():
for strategy, models in zip(strategies, models_list):
models = self.trainer.end_train(models, experiment_name=strategy.name_id)
def routine(
self,
cur_time: Union[str, pd.Timestamp] = None,
@@ -139,33 +145,38 @@ class OnlineManager(Serializable):
cur_time = D.calendar(freq=self.freq).max()
self.cur_time = pd.Timestamp(cur_time) # None for latest date
models_list = []
for strategy in self.strategies:
self.logger.info(f"Strategy `{strategy.name_id}` begins routine...")
if self.status == self.STATUS_NORMAL:
strategy.tool.update_online_pred()
tasks = strategy.prepare_tasks(self.cur_time, **task_kwargs)
models = self.trainer.train(tasks)
if self.status == self.STATUS_NORMAL or not self.trainer.is_delay():
models = self.trainer.end_train(models, experiment_name=strategy.name_id)
models = self.trainer.train(tasks, experiment_name=strategy.name_id)
models_list.append(models)
self.logger.info(f"Finished training {len(models)} models.")
online_models = strategy.prepare_online_models(models, **model_kwargs)
self.history.setdefault(self.cur_time, {})[strategy] = online_models
if not self.trainer.is_delay():
if not self.status == self.STATUS_SIMULATING or not self.trainer.is_delay():
for strategy, models in zip(self.strategies, models_list):
models = self.trainer.end_train(models, experiment_name=strategy.name_id)
self.prepare_signals(**signal_kwargs)
def get_collector(self) -> MergeCollector:
def get_collector(self, **kwargs) -> MergeCollector:
"""
Get the instance of `Collector <../advanced/task_management.html#Task Collecting>`_ to collect results from every strategy.
This collector can be a basis as the signals preparation.
Args:
**kwargs: the params for get_collector.
Returns:
MergeCollector: the collector to merge other collectors.
"""
collector_dict = {}
for strategy in self.strategies:
collector_dict[strategy.name_id] = strategy.get_collector()
collector_dict[strategy.name_id] = strategy.get_collector(**kwargs)
return MergeCollector(collector_dict, process_list=[])
def add_strategy(self, strategies: Union[OnlineStrategy, List[OnlineStrategy]]):
@@ -297,6 +308,7 @@ class OnlineManager(Serializable):
# NOTE: Assumption: the predictions of online models need less than next cur_time, or this method will work in a wrong way.
self.prepare_signals(**signal_kwargs)
if signals_time > cur_time:
# FIXME: if use DelayTrainer and worker (and worker is faster than main progress), there are some possibilities of showing this warning.
self.logger.warn(
f"The signals have already parpred to {signals_time} by last preparation, but current time is only {cur_time}. This may be because the online models predict more than they should, which can cause signals to be contaminated by the offline models."
)

76
qlib/workflow/record_temp.py
View File

@@ -17,7 +17,7 @@ from ..log import get_module_logger
from ..utils import flatten_dict
from ..utils.resam import parse_freq
from ..strategy.base import BaseStrategy
from ..contrib.eva.alpha import calc_ic, calc_long_short_return
from ..contrib.eva.alpha import calc_ic, calc_long_short_return, calc_long_short_prec
logger = get_module_logger("workflow", logging.INFO)
@@ -302,7 +302,7 @@ class PortAnaRecord(RecordTemp):
define the executor class as well as the kwargs.
config["backtest"] : dict
define the backtest kwargs.
risk_analysis_freq : int
risk_analysis_freq : str|List[str]
risk analysis freq of report
"""
super().__init__(recorder=recorder, **kwargs)
@@ -310,8 +310,11 @@ class PortAnaRecord(RecordTemp):
self.strategy_config = config["strategy"]
self.executor_config = config["executor"]
self.backtest_config = config["backtest"]
_count, _freq = parse_freq(risk_analysis_freq)
self.risk_analysis_freq = f"{_count}{_freq}"
if isinstance(risk_analysis_freq, str):
risk_analysis_freq = [risk_analysis_freq]
self.risk_analysis_freq = [
"{0}{1}".format(*parse_freq(_analysis_freq)) for _analysis_freq in risk_analysis_freq
]
self.report_freq = self._get_report_freq(self.executor_config)
def _get_report_freq(self, executor_config):
@@ -336,34 +339,35 @@ class PortAnaRecord(RecordTemp):
**{f"positions_normal_{report_freq}.pkl": positions_normal}, artifact_path=PortAnaRecord.get_path()
)
if self.risk_analysis_freq not in report_dict:
warnings.warn(
f"the freq {self.risk_analysis_freq} report is not found, please set the corresponding env with `generate_report==True`"
)
else:
report_normal, _ = report_dict.get(self.risk_analysis_freq)
analysis = dict()
analysis["excess_return_without_cost"] = risk_analysis(
report_normal["return"] - report_normal["bench"], freq=self.risk_analysis_freq
)
analysis["excess_return_with_cost"] = risk_analysis(
report_normal["return"] - report_normal["bench"] - report_normal["cost"], freq=self.risk_analysis_freq
)
analysis_df = pd.concat(analysis) # type: pd.DataFrame
# log metrics
self.recorder.log_metrics(**flatten_dict(analysis_df["risk"].unstack().T.to_dict()))
# save results
self.recorder.save_objects(
**{f"port_analysis_{report_freq}.pkl": analysis_df}, artifact_path=PortAnaRecord.get_path()
)
logger.info(
f"Portfolio analysis record 'port_analysis_{report_freq}.pkl' has been saved as the artifact of the Experiment {self.recorder.experiment_id}"
)
# print out results
pprint("The following are analysis results of the excess return without cost.")
pprint(analysis["excess_return_without_cost"])
pprint("The following are analysis results of the excess return with cost.")
pprint(analysis["excess_return_with_cost"])
for _analysis_freq in self.risk_analysis_freq:
if _analysis_freq not in report_dict:
warnings.warn(
f"the freq {_analysis_freq} report is not found, please set the corresponding env with `generate_report==True`"
)
else:
report_normal, _ = report_dict.get(_analysis_freq)
analysis = dict()
analysis["excess_return_without_cost"] = risk_analysis(
report_normal["return"] - report_normal["bench"], freq=_analysis_freq
)
analysis["excess_return_with_cost"] = risk_analysis(
report_normal["return"] - report_normal["bench"] - report_normal["cost"], freq=_analysis_freq
)
analysis_df = pd.concat(analysis) # type: pd.DataFrame
# log metrics
self.recorder.log_metrics(**flatten_dict(analysis_df["risk"].unstack().T.to_dict()))
# save results
self.recorder.save_objects(
**{f"port_analysis_{report_freq}.pkl": analysis_df}, artifact_path=PortAnaRecord.get_path()
)
logger.info(
f"Portfolio analysis record 'port_analysis_{report_freq}.pkl' has been saved as the artifact of the Experiment {self.recorder.experiment_id}"
)
# print out results
pprint("The following are analysis results of the excess return without cost.")
pprint(analysis["excess_return_without_cost"])
pprint("The following are analysis results of the excess return with cost.")
pprint(analysis["excess_return_with_cost"])
def list(self):
list_path = []
@@ -374,6 +378,10 @@ class PortAnaRecord(RecordTemp):
PortAnaRecord.get_path(f"positions_normal_{_freq}.pkl"),
]
)
if _freq == self.risk_analysis_freq:
list_path.append(PortAnaRecord.get_path(f"port_analysis_{_freq}.pkl"))
for _analysis_freq in self.risk_analysis_freq:
if _analysis_freq in self.report_freq:
list_path.append(PortAnaRecord.get_path(f"port_analysis_{_analysis_freq}.pkl"))
else:
warnings.warn(f"{_analysis_freq} is not found")
return list_path

61
qlib/workflow/task/manage.py
View File

@@ -69,28 +69,29 @@ class TaskManager:
ENCODE_FIELDS_PREFIX = ["def", "res"]
def __init__(self, task_pool: str = None):
def __init__(self, task_pool: str):
"""
Init Task Manager, remember to make the statement of MongoDB url and database name firstly.
A TaskManager instance serves a specific task pool.
The static method of this module serves the whole MongoDB.
Parameters
----------
task_pool: str
the name of Collection in MongoDB
"""
self.mdb = get_mongodb()
if task_pool is not None:
self.task_pool = getattr(self.mdb, task_pool)
self.task_pool = getattr(get_mongodb(), task_pool)
self.logger = get_module_logger(self.__class__.__name__)
def list(self) -> list:
@staticmethod
def list() -> list:
"""
List the all collection(task_pool) of the db
List the all collection(task_pool) of the db.
Returns:
list
"""
return self.mdb.list_collection_names()
return get_mongodb().list_collection_names()
def _encode_task(self, task):
for prefix in self.ENCODE_FIELDS_PREFIX:
@@ -109,6 +110,25 @@ class TaskManager:
def _dict_to_str(self, flt):
return {k: str(v) for k, v in flt.items()}
def _decode_query(self, query):
"""
If the query includes any `_id`, then it needs `ObjectId` to decode.
For example, when using TrainerRM, it needs query `{"_id": {"$in": _id_list}}`. Then we need to `ObjectId` every `_id` in `_id_list`.
Args:
query (dict): query dict. Defaults to {}.
Returns:
dict: the query after decoding.
"""
if "_id" in query:
if isinstance(query["_id"], dict):
for key in query["_id"]:
query["_id"][key] = [ObjectId(i) for i in query["_id"][key]]
else:
query["_id"] = ObjectId(query["_id"])
return query
def replace_task(self, task, new_task):
"""
Use a new task to replace a old one
@@ -224,8 +244,7 @@ class TaskManager:
dict: a task(document in collection) after decoding
"""
query = query.copy()
if "_id" in query:
query["_id"] = ObjectId(query["_id"])
query = self._decode_query(query)
query.update({"status": status})
task = self.task_pool.find_one_and_update(
query, {"$set": {"status": self.STATUS_RUNNING}}, sort=[("priority", pymongo.DESCENDING)]
@@ -283,12 +302,11 @@ class TaskManager:
dict: a task(document in collection) after decoding
"""
query = query.copy()
if "_id" in query:
query["_id"] = ObjectId(query["_id"])
query = self._decode_query(query)
for t in self.task_pool.find(query):
yield self._decode_task(t)
def re_query(self, _id):
def re_query(self, _id) -> dict:
"""
Use _id to query task.
@@ -339,8 +357,7 @@ class TaskManager:
"""
query = query.copy()
if "_id" in query:
query["_id"] = ObjectId(query["_id"])
query = self._decode_query(query)
self.task_pool.delete_many(query)
def task_stat(self, query={}) -> dict:
@@ -354,8 +371,7 @@ class TaskManager:
dict
"""
query = query.copy()
if "_id" in query:
query["_id"] = ObjectId(query["_id"])
query = self._decode_query(query)
tasks = self.query(query=query, decode=False)
status_stat = {}
for t in tasks:
@@ -377,8 +393,7 @@ class TaskManager:
def reset_status(self, query, status):
query = query.copy()
if "_id" in query:
query["_id"] = ObjectId(query["_id"])
query = self._decode_query(query)
print(self.task_pool.update_many(query, {"$set": {"status": status}}))
def prioritize(self, task, priority: int):
@@ -402,9 +417,19 @@ class TaskManager:
return sum(task_stat.values())
def wait(self, query={}):
"""
When multiprocessing, the main progress may fetch nothing from TaskManager because there are still some running tasks.
So main progress should wait until all tasks are trained well by other progress or machines.
Args:
query (dict, optional): the query dict. Defaults to {}.
"""
task_stat = self.task_stat(query)
total = self._get_total(task_stat)
last_undone_n = self._get_undone_n(task_stat)
if last_undone_n == 0:
return
self.logger.warn(f"Waiting for {last_undone_n} undone tasks. Please make sure they are running.")
with tqdm(total=total, initial=total - last_undone_n) as pbar:
while True:
time.sleep(10)

9
qlib/workflow/utils.py
View File

@@ -17,7 +17,6 @@ def experiment_exit_handler():
Thus, if any exception or user interuption occurs beforehead, we should handle them first. Once `R` is
ended, another call of `R.end_exp` will not take effect.
"""
signal.signal(signal.SIGINT, experiment_kill_signal_handler) # handle user keyboard interupt
sys.excepthook = experiment_exception_hook # handle uncaught exception
atexit.register(R.end_exp, recorder_status=Recorder.STATUS_FI) # will not take effect if experiment ends
@@ -39,11 +38,3 @@ def experiment_exception_hook(type, value, tb):
print(f"{type.__name__}: {value}")
R.end_exp(recorder_status=Recorder.STATUS_FA)
def experiment_kill_signal_handler(signum, frame):
"""
End an experiment when user kill the program through keyboard (CTRL+C, etc.).
"""
R.end_exp(recorder_status=Recorder.STATUS_FA)
raise KeyboardInterrupt