optimize_CI (microsoft#1314)

Author: SunsetWolf

.github/workflows/test_qlib_from_source.yml

@@ -87,9 +87,10 @@ jobs:
       # E1102: not-callable
       # E1136: unsubscriptable-object
     # References for parameters: https://github.com/PyCQA/pylint/issues/4577#issuecomment-1000245962
+    # We use sys.setrecursionlimit(2000) to make the recursion depth larger to ensure that pylint works properly (the default recursion depth is 1000).
     - name: Check Qlib with pylint
       run: |
-        pylint --disable=C0104,C0114,C0115,C0116,C0301,C0302,C0411,C0413,C1802,R0401,R0801,R0902,R0903,R0911,R0912,R0913,R0914,R0915,R1720,W0105,W0123,W0201,W0511,W0613,W1113,W1514,E0401,E1121,C0103,C0209,R0402,R1705,R1710,R1725,R1735,W0102,W0212,W0221,W0223,W0231,W0237,W0612,W0621,W0622,W0703,W1309,E1102,E1136 --const-rgx='[a-z_][a-z0-9_]{2,30}$' qlib --init-hook "import astroid; astroid.context.InferenceContext.max_inferred = 500"
+        pylint --disable=C0104,C0114,C0115,C0116,C0301,C0302,C0411,C0413,C1802,R0401,R0801,R0902,R0903,R0911,R0912,R0913,R0914,R0915,R1720,W0105,W0123,W0201,W0511,W0613,W1113,W1514,E0401,E1121,C0103,C0209,R0402,R1705,R1710,R1725,R1735,W0102,W0212,W0221,W0223,W0231,W0237,W0612,W0621,W0622,W0703,W1309,E1102,E1136 --const-rgx='[a-z_][a-z0-9_]{2,30}$' qlib --init-hook "import astroid; astroid.context.InferenceContext.max_inferred = 500; import sys; sys.setrecursionlimit(2000)"
 
     # The following flake8 error codes were ignored:
       # E501 line too long

examples/rl/scripts/collect_pickle_dataframe.py

@@ -13,7 +13,7 @@
     df = pd.concat(list(df.values())).reset_index()
     df["date"] = df["datetime"].dt.date.astype("datetime64")
     instruments = sorted(set(df["instrument"]))
-    
+
     os.makedirs(os.path.join("data", "pickle_dataframe", tag), exist_ok=True)
     for instrument in tqdm(instruments):
         cur = df[df["instrument"] == instrument].sort_values(by=["datetime"])

examples/rl/scripts/gen_backtest_orders.py

@@ -22,19 +22,21 @@
 df_list = []
 for instrument in instruments:
     print(instrument)
-    
+
     cur_df = df[df["instrument"] == instrument]
-    
+
     dates = sorted(set([str(d).split(" ")[0] for d in cur_df["date"]]))
-    
+
     n = args.num_order
     df_list.append(
-        pd.DataFrame({
-            "date": sorted(np.random.choice(dates, size=n, replace=False)),
-            "instrument": [instrument] * n,
-            "amount": np.random.randint(low=3, high=11, size=n) * 100.0,
-            "order_type": np.random.randint(low=0, high=2, size=n),
-        }).set_index(["date", "instrument"]),
+        pd.DataFrame(
+            {
+                "date": sorted(np.random.choice(dates, size=n, replace=False)),
+                "instrument": [instrument] * n,
+                "amount": np.random.randint(low=3, high=11, size=n) * 100.0,
+                "order_type": np.random.randint(low=0, high=2, size=n),
+            }
+        ).set_index(["date", "instrument"]),
     )
 
 total_df = pd.concat(df_list)

examples/rl/scripts/gen_pickle_data.py

@@ -30,8 +30,8 @@
     if "backtest_conf" in conf:
         backtest = provider._gen_dataframe(deepcopy(provider.backtest_conf))
 
-    provider.feature_conf['path'] = os.path.splitext(provider.feature_conf['path'])[0] + '/'
-    provider.backtest_conf['path'] = os.path.splitext(provider.backtest_conf['path'])[0] + '/'
+    provider.feature_conf["path"] = os.path.splitext(provider.feature_conf["path"])[0] + "/"
+    provider.backtest_conf["path"] = os.path.splitext(provider.backtest_conf["path"])[0] + "/"
     # Split by date
     if args.split == "date" or args.split == "both":
         provider._gen_day_dataset(deepcopy(provider.feature_conf), "feature")

examples/rl/scripts/gen_training_orders.py

@@ -23,15 +23,17 @@
     path = os.path.join("data", "pickle", f"backtest{group}.pkl")
     df = pickle.load(open(path, "rb")).reset_index()
     df["date"] = df["datetime"].dt.date.astype("datetime64")
-    
+
     dates = sorted(set([str(d).split(" ")[0] for d in df["date"]]))
 
-    data_df = pd.DataFrame({
-        "date": sorted(np.random.choice(dates, size=n, replace=False)),
-        "instrument": [args.stock] * n,
-        "amount": np.random.randint(low=3, high=11, size=n) * 100.0,
-        "order_type": [0] * n,
-    }).set_index(["date", "instrument"])
+    data_df = pd.DataFrame(
+        {
+            "date": sorted(np.random.choice(dates, size=n, replace=False)),
+            "instrument": [args.stock] * n,
+            "amount": np.random.randint(low=3, high=11, size=n) * 100.0,
+            "order_type": [0] * n,
+        }
+    ).set_index(["date", "instrument"])
 
     os.makedirs(os.path.join("data", "training_order_split", group), exist_ok=True)
     pickle.dump(data_df, open(os.path.join("data", "training_order_split", group, f"{args.stock}.pkl"), "wb"))

qlib/backtest/decision.py

@@ -579,8 +579,11 @@ def __repr__(self) -> str:
 
 
 class TradeDecisionWithDetails(TradeDecisionWO):
-    """Decision with detail information. Detail information is used to generate execution reports.
     """
+    Decision with detail information.
+    Detail information is used to generate execution reports.
+    """
+
     def __init__(
         self,
         order_list: List[Order],

qlib/rl/contrib/backtest.py

@@ -8,13 +8,14 @@
 import pickle
 from collections import defaultdict
 from pathlib import Path
-from typing import List, Literal, Optional, Tuple, Union
+from typing import List, Optional, Tuple, Union
 
 import numpy as np
 import pandas as pd
 import torch
 from joblib import Parallel, delayed
 
+from qlib.typehint import Literal
 from qlib.backtest import collect_data_loop, get_strategy_executor
 from qlib.backtest.decision import BaseTradeDecision, Order, OrderDir, TradeRangeByTime
 from qlib.backtest.executor import BaseExecutor, NestedExecutor, SimulatorExecutor

setup.py

@@ -142,7 +142,11 @@ def get_version(rel_path: str) -> str:
             "setuptools",
             "black",
             "pylint",
-            "mypy",
+            # Using the latest versions(0.981 and 0.982) of mypy,
+            # the error "multiprocessing.Value()" is detected in the file "qlib/rl/utils/data_queue.py",
+            # If this is fixed in a subsequent version of mypy, then we will revert to the latest version of mypy.
+            # References: https://github.com/python/typeshed/issues/8799
+            "mypy<0.981",
             "flake8",
             "readthedocs_sphinx_ext",
             "cmake",

tests/test_all_pipeline.py

@@ -56,39 +56,8 @@ def train(uri_path: str = None):
         ic = sar.load("ic.pkl")
         ric = sar.load("ric.pkl")
 
-    return pred_score, {"ic": ic, "ric": ric}, rid
-
-
-def train_with_sigana(uri_path: str = None):
-    """train model followed by SigAnaRecord
-
-    Returns
-    -------
-        pred_score: pandas.DataFrame
-            predict scores
-        performance: dict
-            model performance
-    """
-    model = init_instance_by_config(CSI300_GBDT_TASK["model"])
-    dataset = init_instance_by_config(CSI300_GBDT_TASK["dataset"])
-    # start exp
-    with R.start(experiment_name="workflow_with_sigana", uri=uri_path):
-        R.log_params(**flatten_dict(CSI300_GBDT_TASK))
-        model.fit(dataset)
-        recorder = R.get_recorder()
-
-        sr = SignalRecord(model, dataset, recorder)
-        sr.generate()
-        pred_score = sr.load("pred.pkl")
-
-        # predict and calculate ic and ric
-        sar = SigAnaRecord(recorder)
-        sar.generate()
-        ic = sar.load("ic.pkl")
-        ric = sar.load("ric.pkl")
-
         uri_path = R.get_uri()
-    return pred_score, {"ic": ic, "ric": ric}, uri_path
+    return pred_score, {"ic": ic, "ric": ric}, rid, uri_path
 
 
 def fake_experiment():
@@ -186,19 +155,13 @@ def tearDownClass(cls) -> None:
         shutil.rmtree(cls.URI_PATH.lstrip("file:"))
 
     @pytest.mark.slow
-    def test_0_train_with_sigana(self):
-        TestAllFlow.PRED_SCORE, ic_ric, uri_path = train_with_sigana(self.URI_PATH)
-        self.assertGreaterEqual(ic_ric["ic"].all(), 0, "train failed")
-        self.assertGreaterEqual(ic_ric["ric"].all(), 0, "train failed")
-
-    @pytest.mark.slow
-    def test_1_train(self):
-        TestAllFlow.PRED_SCORE, ic_ric, TestAllFlow.RID = train(self.URI_PATH)
+    def test_0_train(self):
+        TestAllFlow.PRED_SCORE, ic_ric, TestAllFlow.RID, uri_path = train(self.URI_PATH)
         self.assertGreaterEqual(ic_ric["ic"].all(), 0, "train failed")
         self.assertGreaterEqual(ic_ric["ric"].all(), 0, "train failed")
 
     @pytest.mark.slow
-    def test_2_backtest(self):
+    def test_1_backtest(self):
         analyze_df = backtest_analysis(TestAllFlow.PRED_SCORE, TestAllFlow.RID, self.URI_PATH)
         self.assertGreaterEqual(
             analyze_df.loc(axis=0)["excess_return_with_cost", "annualized_return"].values[0],
@@ -208,7 +171,7 @@ def test_2_backtest(self):
         self.assertTrue(not analyze_df.isna().any().any(), "backtest failed")
 
     @pytest.mark.slow
-    def test_3_expmanager(self):
+    def test_2_expmanager(self):
         pass_default, pass_current, uri_path = fake_experiment()
         self.assertTrue(pass_default, msg="default uri is incorrect")
         self.assertTrue(pass_current, msg="current uri is incorrect")
@@ -217,10 +180,9 @@ def test_3_expmanager(self):
 
 def suite():
     _suite = unittest.TestSuite()
-    _suite.addTest(TestAllFlow("test_0_train_with_sigana"))
-    _suite.addTest(TestAllFlow("test_1_train"))
-    _suite.addTest(TestAllFlow("test_2_backtest"))
-    _suite.addTest(TestAllFlow("test_3_expmanager"))
+    _suite.addTest(TestAllFlow("test_0_train"))
+    _suite.addTest(TestAllFlow("test_1_backtest"))
+    _suite.addTest(TestAllFlow("test_2_expmanager"))
     return _suite
 
 

tests/test_contrib_workflow.py

@@ -11,7 +11,24 @@
 from qlib.utils import init_instance_by_config, flatten_dict
 from qlib.workflow import R
 from qlib.tests import TestAutoData
-from qlib.tests.config import CSI300_GBDT_TASK
+from qlib.tests.config import GBDT_MODEL, get_dataset_config, CSI300_MARKET
+
+
+CSI300_GBDT_TASK = {
+    "model": GBDT_MODEL,
+    "dataset": get_dataset_config(
+        train=("2020-05-01", "2020-06-01"),
+        valid=("2020-06-01", "2020-07-01"),
+        test=("2020-07-01", "2020-08-01"),
+        handler_kwargs={
+            "start_time": "2020-05-01",
+            "end_time": "2020-08-01",
+            "fit_start_time": "<dataset.kwargs.segments.train.0>",
+            "fit_end_time": "<dataset.kwargs.segments.train.1>",
+            "instruments": CSI300_MARKET,
+        },
+    ),
+}
 
 
 def train_multiseg(uri_path: str = None):