feat(Baseline): added one-split baseline, pure sklearn (#247)

* feat(Baseline): added sklearn_baseline

* feat(Baseline): completely working

config/hashing.py

@@ -20,7 +20,6 @@ def hash_data_config(data_config: dict) -> str:
                 data_config["target_asset"].path
                 + data_config["target_asset"].file_name,
                 data_config["load_non_target_asset"],
-                data_config["start_date"],
                 hash_feature_extractors(data_config["own_features"]),
                 hash_feature_extractors(data_config["other_features"]),
                 hash_feature_extractors(data_config["exogenous_features"]),

config/presets.py

@@ -54,6 +54,7 @@ def get_minimal_config() -> RawConfig:
     meta_models = ["LogisticRegression_two_class", "LGBM"]
 
     return RawConfig(
+        start_date="2021-01-01",
         dimensionality_reduction_ratio=0,
         n_features_to_select=0,
         initial_window_size=3800,
@@ -71,7 +72,7 @@ def get_minimal_config() -> RawConfig:
         meta_models=meta_models,
         event_filter="none",
         event_filter_multiplier=3.5,
-        remove_overlapping_events=True,
+        remove_overlapping_events=False,
         labeling="two_class",
         forecasting_horizon=10,
         transaction_costs=0.002,

data_loader/load.py

@@ -11,20 +11,20 @@ from config.hashing import hash_data_config
 from .types import XDataFrame, ReturnSeries
 from diskcache import Cache
 
-cache = Cache(".cachedir/data")
+# cache = Cache(".cachedir/data")
 
 
-def load_data(**kwargs) -> tuple[XDataFrame, ReturnSeries]:
-    hashed = hash_data_config(kwargs)
-    if hashed in cache:
-        return cache.get(hashed)
-    else:
-        return_value = __load_data(**kwargs)
-        cache[hashed] = return_value
-        return return_value
+# def load_data(**kwargs) -> tuple[XDataFrame, ReturnSeries]:
+#     hashed = hash_data_config(kwargs)
+#     if hashed in cache:
+#         return cache.get(hashed)
+#     else:
+#         return_value = __load_data(**kwargs)
+#         cache[hashed] = return_value
+#         return return_value
 
 
-def __load_data(
+def load_data(
     assets: DataCollection,
     other_assets: DataCollection,
     exogenous_data: DataCollection,

run_baseline.py

@@ -0,0 +1,69 @@
+from re import S
+from typing import Optional
+
+from sklearn.model_selection import train_test_split
+
+from config.types import Config, RawConfig
+from config.preprocess import preprocess_config
+from config.presets import get_default_config, get_minimal_config
+
+from data_loader.load import load_data
+from data_loader.process import check_data
+
+from labeling.process import label_data
+
+from sklearn.metrics import accuracy_score
+
+
+def run_sklearn_pipeline(raw_config: RawConfig):
+    config = preprocess_config(raw_config)
+    run_sklearn_training(config)
+
+
+def run_sklearn_training(config: Config):
+
+    print("---> Load data, check for validity")
+    X, returns = load_data(
+        assets=config.assets,
+        other_assets=config.other_assets,
+        exogenous_data=config.exogenous_data,
+        target_asset=config.target_asset,
+        load_non_target_asset=config.load_non_target_asset,
+        own_features=config.own_features,
+        other_features=config.other_features,
+        exogenous_features=config.exogenous_features,
+        start_date=config.start_date,
+    )
+
+    assert check_data(X, config) == True, "Data is not valid."
+
+    print("---> Filter for significant events when we want to trade, and label data")
+    events, X, y, forward_returns = label_data(
+        event_filter=config.event_filter,
+        event_labeller=config.labeling,
+        X=X,
+        returns=returns,
+        remove_overlapping_events=config.remove_overlapping_events,
+    )
+
+    X_train, X_test, y_train, y_test = train_test_split(
+        X, y, test_size=0.2, random_state=42, shuffle=False
+    )
+
+    print("---> Train directional models")
+
+    for transformation in config.transformations:
+        X_train = transformation.fit_transform(X_train, y_train)
+
+    for transformation in config.transformations:
+        X_test = transformation.transform(X_test)
+
+    config.directional_model.fit(X_train, y_train)
+    preds = config.directional_model.predict(X_test)
+    print(accuracy_score(y_test, preds))
+
+
+if __name__ == "__main__":
+    run_sklearn_pipeline(
+        raw_config=get_minimal_config(),
+    )