Merge pull request #33 from HippolyteGuigon/features_building_dockerf…

…iles

pushing_final_tests

Random_forest/model/model.py

@@ -298,4 +298,27 @@ def predict(self, full_X_to_predict: np.array) -> np.array:
             prediction = self.individual_predict(x)
             predicted.append(prediction)
 
+        predicted = np.array(predicted)
+
         return predicted
+
+    def score(self, y_pred: np.array, y_test: np.array) -> float:
+        """
+        The goal of this function is, once
+        the model has been fitted, to compute
+        its score regarding its objective (regression
+        or classification)
+
+        Arguments:
+            -y_pred: np.array: The predictions made
+            by the model
+            -y_test: np.array: The real values
+        """
+
+        if self.objective == "classification":
+            return np.mean(y_pred == y_test)
+        else:
+            rss = ((y_pred - y_test) ** 2).sum()
+            tss = ((y_test - y_test.mean()) ** 2).sum()
+            r_squared = 1 - rss / tss
+            return r_squared

main.py

@@ -3,7 +3,7 @@
 from sklearn.metrics import mean_absolute_error
 from sklearn.model_selection import train_test_split
 from Random_forest.model.model import RandomForest
-from Random_forest.decision_tree.array_functions import get_random_set, is_float
+from Random_forest.decision_tree.array_functions import get_random_set
 from Random_forest.configs.confs import load_conf
 
 main_params = load_conf("configs/main.yml", include=True)

test_unnitest.py

@@ -195,6 +195,53 @@ def test_model_prediction(self) -> None:
         self.assertTrue(np.all(classification_predictions))
         self.assertTrue(np.all(regression_predictions))
 
+    def test_score(self) -> None:
+        """
+        The goal of this function
+        is to test the model ability
+        to score a result
+
+        Arguments:
+            -None
+
+        Returns:
+            -None
+        """
+
+        X, y = get_random_set(
+            row_size_test_dataset=row_size_test_dataset, objective="classification"
+        )
+        X_predict_classification, y_predict_classification = get_random_set(
+            row_size_test_dataset=30, objective="classification"
+        )
+
+        model_classification = RandomForest(objective="classification")
+        model_classification.fit(X, y)
+
+        classification_predictions = model_classification.predict(
+            X_predict_classification
+        )
+
+        X, y = get_random_set(
+            row_size_test_dataset=row_size_test_dataset, objective="regression"
+        )
+        X_predict_regression, y_predict_regression = get_random_set(
+            row_size_test_dataset=30, objective="regression"
+        )
+
+        model_regression = RandomForest(objective="regression")
+        model_regression.fit(X, y)
+
+        regression_predictions = model_regression.predict(X_predict_regression)
+
+        accuracy = model_classification.score(
+            classification_predictions, y_predict_classification
+        )
+        r_squared = model_regression.score(regression_predictions, y_predict_regression)
+
+        self.assertTrue((accuracy >= 0 and accuracy <= 1))
+        self.assertTrue(isinstance(r_squared, float))
+
 
 if __name__ == "__main__":
     main()