feat(Models): added StaticAverageModel for average ensembling & StaticNaiveModel (#64)

* feat(Models): added StaticAverageModel for average ensembling

* feat(Models): made sure we only pipe in predictions to StaticAverageModel, added StaticNaiveModel as potential baseline

* chore(Models): removed unnecessary commented out code

models/average.py

@@ -0,0 +1,24 @@
+from models.base import Model
+import numpy as np
+
+class StaticAverageModel(Model):
+    '''
+    Model that averages .
+    '''
+
+    # data_format = 'dataframe'
+    data_scaling = 'unscaled'
+    only_column = 'model_'
+
+    def fit(self, X, y):
+        # This is a static model, it can' learn anything
+        pass
+
+    def predict(self, X):
+        # Make sure there's data to average
+        assert X.shape[1] > 0
+        prediction = np.average(X[0])
+        return np.array([prediction])
+
+    def clone(self):
+        return self

models/base.py

@@ -1,20 +1,24 @@
 
 from typing import Literal, Optional
 from sklearn.base import clone
+from abc import ABC, abstractmethod, abstractproperty
 
-class Model:
+class Model(ABC):
 
     # data_format: Literal['dataframe', 'numpy'] 
     data_scaling: Literal["scaled", "unscaled"]
     # data_format: Literal["wide", "narrow"]
     only_column: Optional[str]
 
+    @abstractmethod
     def fit(self, X, y):
         pass
 
+    @abstractmethod
     def predict(self, X):
         pass
 
+    @abstractmethod
     def clone(self):
         pass
 

models/momentum.py

@@ -6,7 +6,7 @@ class StaticMomentumModel(Model):
     Model that uses only one feature: momentum. It's positive if momentum is greater than 0, otherwise it's negative.
     '''
 
-    data_format = 'dataframe'
+    # data_format = 'dataframe'
     data_scaling = 'unscaled'
     only_column = 'mom'
 

models/naive.py

@@ -0,0 +1,21 @@
+from models.base import Model
+import numpy as np
+
+class StaticNaiveModel(Model):
+    '''
+    Model that carries the last observation (from returns) to the next one, naively.
+    '''
+
+    # data_format = 'dataframe'
+    data_scaling = 'unscaled'
+    only_column = None
+
+    def fit(self, X, y):
+        # This is a static model, it can' learn anything
+        pass
+
+    def predict(self, X):
+        return np.array([X[-1][0]])
+
+    def clone(self):
+        return self

run_pipeline.py

@@ -15,6 +15,8 @@ from sklearn.neural_network import MLPRegressor, MLPClassifier
 from sklearn.ensemble import AdaBoostRegressor, RandomForestRegressor, ExtraTreesRegressor, AdaBoostClassifier, GradientBoostingClassifier, RandomForestClassifier, ExtraTreesClassifier
 from models.base import SKLearnModel
 from models.momentum import StaticMomentumModel
+from models.average import StaticAverageModel
+from models.naive import StaticNaiveModel
 
 import feature_extractors.feature_extractor_presets as feature_extractor_presets
 from training.pipeline import run_single_asset_trainig_pipeline
@@ -53,19 +55,22 @@ def get_config() -> tuple[dict, dict, dict]:
         # ('RF', SKLearnModel(RandomForestRegressor(n_jobs=-1))),
         # ('SVR', SKLearnModel(SVR(kernel='rbf', C=1e3, gamma=0.1)))
     ]
-    regression_ensemble_model = [('Ensemble - Ridge', SKLearnModel(Ridge(alpha=0.1)))]
+    regression_ensemble_model = [('Ensemble - Average', StaticAverageModel())]
+    # regression_ensemble_model = [('Ensemble - Ridge', SKLearnModel(Ridge(alpha=0.1)))]
 
     classification_models = [
         ('LR', SKLearnModel(LogisticRegression(n_jobs=-1))),
-        # ('LDA', SKLearnModel(LinearDiscriminantAnalysis())),
+        ('LDA', SKLearnModel(LinearDiscriminantAnalysis())),
-        # ('KNN', SKLearnModel(KNeighborsClassifier())),
+        ('KNN', SKLearnModel(KNeighborsClassifier())),
-        # ('CART', SKLearnModel(DecisionTreeClassifier())),
+        ('CART', SKLearnModel(DecisionTreeClassifier())),
-        ('StaticMomentum', StaticMomentumModel(allow_short=True))
+        ('StaticMomentum', StaticMomentumModel(allow_short=True)),
+        # ('StaticNaive', StaticNaiveModel()),
         # ('NB', SKLearnModel(GaussianNB())),
         # ('AB', SKLearnModel(AdaBoostClassifier())),
         # ('RF', SKLearnModel(RandomForestClassifier(n_jobs=-1)))
     ]
-    classification_ensemble_model = [('Ensemble - CART', SKLearnModel(DecisionTreeClassifier()))]
+    classification_ensemble_model = [('Ensemble - Average', StaticAverageModel())]
+    # classification_ensemble_model = [('Ensemble - CART', SKLearnModel(DecisionTreeClassifier()))]
 
     model_config = dict(
         level_1_models = regression_models if data_config['method'] == 'regression' else classification_models,

training/pipeline.py

@@ -56,7 +56,8 @@ def run_single_asset_trainig_pipeline(
         )
         column_name = ticker_to_predict + "_" + model_name
         results[column_name] = result
-        predictions[column_name] = preds
+        # column names for model outputs should be different, so we can differentiate between original data and model predictions later, where necessary
+        predictions["model_" + column_name] = preds
         
         if wandb_active:
             run = wandb.init(project="price-forecasting", config={"model_type": model_name, "ticker": ticker_to_predict}, reinit=True)