handler demo cache (#606)

* handler demo  cache

* Update data_cache_demo.py

* example to reusing processed data in memory

* Skip dumping task of task_train

* FIX Black

Co-authored-by: Wangwuyi123 <51237097+Wangwuyi123@users.noreply.github.com>

examples/data_demo/README.md

@@ -0,0 +1,2 @@
+# Introduction
+The examples in this folder try to demonstrate some common usage of data-related modules of Qlib

examples/data_demo/data_cache_demo.py

@@ -0,0 +1,53 @@
+# Copyright (c) Microsoft Corporation.
+# Licensed under the MIT License.
+"""
+    The motivation of this demo
+    - To show the data modules of Qlib is Serializable, users can dump processed data to disk to avoid duplicated data preprocessing
+"""
+
+from copy import deepcopy
+from pathlib import Path
+import pickle
+from pprint import pprint
+import subprocess
+import yaml
+from qlib.log import TimeInspector
+
+from qlib import init
+from qlib.data.dataset.handler import DataHandlerLP
+from qlib.utils import init_instance_by_config
+
+# For general purpose, we use relative path
+DIRNAME = Path(__file__).absolute().resolve().parent
+
+if __name__ == "__main__":
+    init()
+
+    config_path = DIRNAME.parent / "benchmarks/LightGBM/workflow_config_lightgbm_Alpha158.yaml"
+
+    # 1) show original time
+    with TimeInspector.logt("The original time without handler cache:"):
+        subprocess.run(f"qrun {config_path}", shell=True)
+
+    # 2) dump handler
+    task_config = yaml.safe_load(config_path.open())
+    hd_conf = task_config["task"]["dataset"]["kwargs"]["handler"]
+    pprint(hd_conf)
+    hd: DataHandlerLP = init_instance_by_config(hd_conf)
+    hd_path = DIRNAME / "handler.pkl"
+    hd.to_pickle(hd_path, dump_all=True)
+
+    # 3) create new task with handler cache
+    new_task_config = deepcopy(task_config)
+    new_task_config["task"]["dataset"]["kwargs"]["handler"] = f"file://{hd_path}"
+    new_task_config
+    new_task_path = DIRNAME / "new_task.yaml"
+    print("The location of the new task", new_task_path)
+
+    # save new task
+    with new_task_path.open("w") as f:
+        yaml.safe_dump(new_task_config, f)
+
+    # 4) train model with new task
+    with TimeInspector.logt("The time for task with handler cache:"):
+        subprocess.run(f"qrun {new_task_path}", shell=True)

examples/data_demo/data_mem_resuse_demo.py

@@ -0,0 +1,59 @@
+# Copyright (c) Microsoft Corporation.
+# Licensed under the MIT License.
+"""
+    The motivation of this demo
+    - To show the data modules of Qlib is Serializable, users can dump processed data to disk to avoid duplicated data preprocessing
+"""
+
+from copy import deepcopy
+from pathlib import Path
+import pickle
+from pprint import pprint
+import subprocess
+
+import yaml
+
+from qlib import init
+from qlib.data.dataset.handler import DataHandlerLP
+from qlib.log import TimeInspector
+from qlib.model.trainer import task_train
+from qlib.utils import init_instance_by_config
+
+# For general purpose, we use relative path
+DIRNAME = Path(__file__).absolute().resolve().parent
+
+if __name__ == "__main__":
+    init()
+
+    repeat = 2
+    exp_name = "data_mem_reuse_demo"
+
+    config_path = DIRNAME.parent / "benchmarks/LightGBM/workflow_config_lightgbm_Alpha158.yaml"
+    task_config = yaml.safe_load(config_path.open())
+
+    # 1) without using processed data in memory
+    with TimeInspector.logt("The original time without reusing processed data in memory:"):
+        for i in range(repeat):
+            task_train(task_config["task"], experiment_name=exp_name)
+
+    # 2) prepare processed data in memory.
+    hd_conf = task_config["task"]["dataset"]["kwargs"]["handler"]
+    pprint(hd_conf)
+    hd: DataHandlerLP = init_instance_by_config(hd_conf)
+
+    # 3) with reusing processed data in memory
+    new_task = deepcopy(task_config["task"])
+    new_task["dataset"]["kwargs"]["handler"] = hd
+    print(new_task)
+
+    with TimeInspector.logt("The time with reusing processed data in memory:"):
+        # this will save the time to reload and process data from disk(in `DataHandlerLP`)
+        # It still takes a lot of time in the backtest phase
+        for i in range(repeat):
+            task_train(new_task, experiment_name=exp_name)
+
+    # 4) User can change other parts exclude processed data in memory(handler)
+    new_task = deepcopy(task_config["task"])
+    new_task["dataset"]["kwargs"]["segments"]["train"] = ("20100101", "20131231")
+    with TimeInspector.logt("The time with reusing processed data in memory:"):
+        task_train(new_task, experiment_name=exp_name)