liuyu/qlib
1
0
Fork 0
mirror of https://github.com/microsoft/qlib.git synced 2026-07-01 18:11:18 +08:00
Code Activity

update trainer and README.md

Browse Source
This commit is contained in:
Young
2020-11-26 12:40:50 +00:00
parent 7d092f39c8
commit 0fbf401c98
4 changed files with 47 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
README.md
View File

@@ -45,13 +45,11 @@ For more details, please refer to our paper ["Qlib: An AI-oriented Quantitative
At the module level, Qlib is a platform that consists of the above components. The components are designed as loose-coupled modules and each component could be used stand-alone.
| Name | Description |
| ------ | ----- |
| `Data layer` | `DataServer` focuses on providing high-performance infrastructure for users to manage and retrieve raw data. `DataEnhancement` will preprocess the data and provide the best dataset to be fed into the models. |
| `Interday Model` | `Interday model` focuses on producing prediction scores (aka. _alpha_). Models are trained by `Model Creator` and managed by `Model Manager`. Users could choose one or multiple models for prediction. Multiple models could be combined with `Ensemble` module. |
| `Interday Strategy` | `Portfolio Generator` will take prediction scores as input and output the orders based on the current position to achieve the target portfolio. |
| `Intraday Trading` | `Order Executor` is responsible for executing orders output by `Interday Strategy` and returning the executed results. |
| `Analysis` | Users could get a detailed analysis report of forecasting signals and portfolios in this part. |
| Name | Description |
| ------ | ----- |
| `Infrastructure` layer | `Infrastructure` layer provides underlying support for Quant research. `DataServer` provides high-performance infrastructure for users to manage and retrieve raw data. `Trainer` provides flexible interface to control the training process of models which enable algorithms controlling the training process. |
| `Workflow` layer | `Workflow` layer covers the whole workflow of quantitative investment. `Information Extractor` extracts data for models. `Forecast Model` focuses on producing all kinds of forecast signals (e.g. _alpha_, risk) for other modules. With these signals `Portfolio Generator` will generate the target portfolio and produce orders to be executed by `Order Executor`. |
| `Interface` layer | `Interface` layer tries to present a user-friendly interface for the underlying system. `Analyser` module will provide users detailed analysis reports of forecasting signals, portfolios and execution results |
* The modules with hand-drawn style are under development and will be released in the future.
* The modules with dashed borders are highly user-customizable and extendible.

BIN
docs/_static/img/framework.png vendored
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 205 KiB

After

Width:  |  Height:  |  Size: 271 KiB

40
qlib/model/trainer.py Normal file
View File

@@ -0,0 +1,40 @@
# Copyright (c) Microsoft Corporation.
# Licensed under the MIT License.
from qlib.utils import init_instance_by_config, flatten_dict
from qlib.workflow import R
from qlib.workflow.record_temp import SignalRecord
def task_train(config: dict, experiment_name):
"""
task based training
Parameters
----------
config : dict
A dict describing the training process
"""
# model initiaiton
model = init_instance_by_config(config.get("task")["model"])
dataset = init_instance_by_config(config.get("task")["dataset"])
# start exp
with R.start(experiment_name=experiment_name):
# train model
R.log_params(**flatten_dict(config.get("task")))
model.fit(dataset)
recorder = R.get_recorder()
# generate records: prediction, backtest, and analysis
for record in config.get("task")["record"]:
if record["class"] == SignalRecord.__name__:
srconf = {"model": model, "dataset": dataset, "recorder": recorder}
record["kwargs"].update(srconf)
sr = init_instance_by_config(record)
sr.generate()
else:
rconf = {"recorder": recorder}
record["kwargs"].update(rconf)
ar = init_instance_by_config(record)
ar.generate()

28
qlib/workflow/cli.py
View File

@@ -8,9 +8,7 @@ import qlib
import fire
import pandas as pd
import ruamel.yaml as yaml
from qlib.utils import init_instance_by_config, flatten_dict
from qlib.workflow import R
from qlib.workflow.record_temp import SignalRecord
from ..model.trainer import task_train
def get_path_list(path):
@@ -54,29 +52,7 @@ def workflow(config_path, experiment_name="workflow"):
region = config.get("region")
qlib.init(provider_uri=provider_uri, region=region)
# model initiaiton
model = init_instance_by_config(config.get("task")["model"])
dataset = init_instance_by_config(config.get("task")["dataset"])
# start exp
with R.start(experiment_name=experiment_name):
# train model
R.log_params(**flatten_dict(config.get("task")))
model.fit(dataset)
recorder = R.get_recorder()
# generate records: prediction, backtest, and analysis
for record in config.get("task")["record"]:
if record["class"] == SignalRecord.__name__:
srconf = {"model": model, "dataset": dataset, "recorder": recorder}
record["kwargs"].update(srconf)
sr = init_instance_by_config(record)
sr.generate()
else:
rconf = {"recorder": recorder}
record["kwargs"].update(rconf)
ar = init_instance_by_config(record)
ar.generate()
task_train(config, experiment_name=experiment_name)
# function to run worklflow by config