From 1bce1ef2e75f60860f0421b7a64ac524b805819d Mon Sep 17 00:00:00 2001
From: MattVerlynde <matthieu.verlynde@agroparistech.fr>
Date: Fri, 22 Dec 2023 10:04:17 +0100
Subject: [PATCH] UP my solution

---
 sklearn_questions.py | 50 +++++++++++++++++++++++++++++++++++++-------
 1 file changed, 42 insertions(+), 8 deletions(-)

diff --git a/sklearn_questions.py b/sklearn_questions.py
index fa02e0d..2aab4d2 100644
--- a/sklearn_questions.py
+++ b/sklearn_questions.py
@@ -48,7 +48,8 @@
 to compute distances between 2 sets of samples.
 """
 import numpy as np
-import pandas as pd
+# import pandas as pd
+from pandas.api.types import is_datetime64_dtype as is_datetime64
 
 from sklearn.base import BaseEstimator
 from sklearn.base import ClassifierMixin
@@ -82,6 +83,13 @@ 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.X_, self.y_ = X, y
+        self.classes_ = np.unique(self.y_)
+        self.n_features_in_ = X.shape[1]
+
         return self
 
     def predict(self, X):
@@ -97,7 +105,17 @@ 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])
+        check_is_fitted(self)
+        X = check_array(X)
+
+        y_pred = np.zeros(X.shape[0]).astype(self.y_.dtype)
+        dist_matrix = pairwise_distances(X, self.X_, metric='euclidean')
+
+        for i in range(dist_matrix.shape[0]):
+            yi_near = (self.y_[np.argsort(dist_matrix[i])[:self.n_neighbors]])
+            near_unique, counts_unique = np.unique(yi_near, return_counts=True)
+            y_pred[i] = near_unique[np.argmax(counts_unique)]
+
         return y_pred
 
     def score(self, X, y):
@@ -115,7 +133,8 @@ def score(self, X, y):
         score : float
             Accuracy of the model computed for the (X, y) pairs.
         """
-        return 0.
+        s = np.where(self.predict(X) == y, 1, 0).mean()
+        return s
 
 
 class MonthlySplit(BaseCrossValidator):
@@ -155,7 +174,16 @@ def get_n_splits(self, X, y=None, groups=None):
         n_splits : int
             The number of splits.
         """
-        return 0
+        if self.time_col == 'index':
+            X = X.reset_index()
+        self.X_time = X[self.time_col]
+
+        if not is_datetime64(self.X_time):
+            raise ValueError(
+                "The column {} is not a datetime.".format(self.time_col)
+                )
+
+        return len(self.X_time.dt.to_period('M').unique())-1
 
     def split(self, X, y, groups=None):
         """Generate indices to split data into training and test set.
@@ -177,12 +205,18 @@ def split(self, X, y, groups=None):
         idx_test : ndarray
             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()
+
+        index_sorted = X.resample('M',
+                                  on=self.time_col
+                                  ).apply(lambda x: x.index)
+        index_sorted.index = index_sorted.index.to_period('M')
+
         for i in range(n_splits):
-            idx_train = range(n_samples)
-            idx_test = range(n_samples)
+            idx_train = np.array(index_sorted[index_sorted.index[i]])
+            idx_test = np.array(index_sorted[index_sorted.index[i+1]])
             yield (
                 idx_train, idx_test
             )