UP my solution

sklearn_questions.py

@@ -82,6 +82,12 @@ def fit(self, X, y):
         self : instance of KNearestNeighbors
             The current instance of the classifier
         """
+        X, y = check_X_y(X, y)
+        check_classification_targets(y)
+        self.n_features_in_ = X.shape[1]
+        self.classes_ = np.unique(y)
+        self.X_ = X
+        self.y_ = y
         return self
 
     def predict(self, X):
@@ -97,8 +103,20 @@ def predict(self, X):
         y : ndarray, shape (n_test_samples,)
             Predicted class labels for each test data sample.
         """
-        y_pred = np.zeros(X.shape[0])
-        return y_pred
+        check_is_fitted(self)
+        X = check_array(X)
+        n_test_samples = X.shape[0]
+
+        y_pred = np.zeros(n_test_samples)
+
+        distance = pairwise_distances(X, self.X_, metric="euclidean")
+        distance_sorted = np.argsort(distance)[:, : self.n_neighbors]
+        k_classes = self.y_[distance_sorted]
+        y_pred = [
+            max(set(k.tolist()), key=k.tolist().count) for k in k_classes
+        ]
+
+        return np.array(y_pred)
 
     def score(self, X, y):
         """Calculate the score of the prediction.
@@ -115,7 +133,11 @@ def score(self, X, y):
         score : float
             Accuracy of the model computed for the (X, y) pairs.
         """
-        return 0.
+        check_is_fitted(self, ["X_", "y_"])
+        X = check_array(X)
+        y_pred = self.predict(X)
+        score = np.mean(y_pred == y)
+        return score
 
 
 class MonthlySplit(BaseCrossValidator):
@@ -155,7 +177,13 @@ def get_n_splits(self, X, y=None, groups=None):
         n_splits : int
             The number of splits.
         """
-        return 0
+        X = X.reset_index()
+
+        if not isinstance(X[self.time_col][0], pd.Timestamp):
+            raise ValueError("X[time_col] does not contain datetime")
+
+        n_splits = X[self.time_col].dt.to_period("M").nunique() - 1
+        return n_splits
 
     def split(self, X, y, groups=None):
         """Generate indices to split data into training and test set.
@@ -178,11 +206,19 @@ def split(self, X, y, groups=None):
             The testing set indices for that split.
         """
 
-        n_samples = X.shape[0]
         n_splits = self.get_n_splits(X, y, groups)
+        X = X.reset_index()
+        date_col = X[self.time_col]
+        period = X.resample("M", on=self.time_col).count().sort_index().index
+        period_m_y = period.map(lambda x: (x.month, x.year))
         for i in range(n_splits):
-            idx_train = range(n_samples)
-            idx_test = range(n_samples)
-            yield (
-                idx_train, idx_test
-            )
+            df_train = X[
+                (date_col.dt.month == period_m_y[i][0])
+                & (date_col.dt.year == period_m_y[i][1])
+            ]
+            df_test = X[
+                (date_col.dt.month == period_m_y[i + 1][0])
+                & (date_col.dt.year == period_m_y[i + 1][1])
+            ]
+
+            yield (df_train.index.to_numpy(), df_test.index.to_numpy())