simon_VITAL assignment

.vscode/settings.json

@@ -0,0 +1,4 @@
+{
+    "black-formatter.args": ["--line-length", "79"],
+    "editor.rulers": [79]
+}

numpy_questions.py

@@ -15,6 +15,7 @@
 This will be enforced with `flake8`. You can check that there is no flake8
 errors by calling `flake8` at the root of the repo.
 """
+
 import numpy as np
 
 
@@ -37,12 +38,22 @@ def max_index(X):
         If the input is not a numpy array or
         if the shape is not 2D.
     """
+    if X is None:
+        raise ValueError
+    if not isinstance(X, np.ndarray):
+        raise ValueError
+    if len(X.shape) != 2:
+        raise ValueError
+
     i = 0
     j = 0
 
-    # TODO
+    max = np.max(X)
+    (i, j), *_ = np.argwhere(X == max)
+    return (i, j)
 
-    return i, j
+
+# print(max_index(np.array([[0, 1], [2, 0]])))
 
 
 def wallis_product(n_terms):
@@ -64,4 +75,19 @@ def wallis_product(n_terms):
     """
     # XXX : The n_terms is an int that corresponds to the number of
     # terms in the product. For example 10000.
-    return 0.
+
+    if n_terms == 0:
+        return 1
+
+    indices = np.arange(1, n_terms + 1)
+    x = 4 * indices**2
+    factors = x / (x - 1)
+    prod = 2 * np.prod(factors)
+    return prod
+
+
+# print(wallis_product(0))
+# print(wallis_product(1))
+# print(wallis_product(10))
+# print(wallis_product(10000))
+# print(wallis_product(20000))

sklearn_questions.py

@@ -19,6 +19,7 @@
 for the methods you code and for the class. The docstring will be checked using
 `pydocstyle` that you can also call at the root of the repo.
 """
+
 import numpy as np
 from sklearn.base import BaseEstimator
 from sklearn.base import ClassifierMixin
@@ -30,46 +31,120 @@
 
 
 class OneNearestNeighbor(ClassifierMixin, BaseEstimator):
-    "OneNearestNeighbor classifier."
+    """OneNearestNeighbor classifier."""
 
     def __init__(self):  # noqa: D107
         pass
 
     def fit(self, X, y):
-        """Write docstring.
+        """Fit the model. But for KNN (Here 1NN) there is no real fit to do.
+
+        Parameters
+        ----------
+        X : ndarray of shape (n_samples, n_features)
+                The input array.
+
+        y : ndarray of shape (n_samples,)
+                The output array.
 
-        And describe parameters
+        Returns
+        -------
+        self
         """
+        X, y = validate_data(self, X, y)
+
         X, y = check_X_y(X, y)
         check_classification_targets(y)
         self.classes_ = np.unique(y)
         self.n_features_in_ = X.shape[1]
+        self.X_train_ = X
+        self.y_train_ = y
 
         # XXX fix
         return self
 
     def predict(self, X):
-        """Write docstring.
-
-        And describe parameters
+        """Return the class predicted.
+
+        Parameters
+        ----------
+        X : ndarray of shape (n_samples, n_features)
+            The input array.
+
+        Returns
+        -------
+        y_pred : ndarray of shape (n_samples,)
+                    The output array.
+
+        Raises
+        ------
+        ValueError
+            if the shape is not 2D.
         """
         check_is_fitted(self)
+        X = validate_data(self, X, reset=False)
+
+        if len(X.shape) != 2:
+            raise ValueError
+
         X = check_array(X)
+        # if self.n_features_in_ != X.shape[1]:
+        #     raise ValueError
+
         y_pred = np.full(
-            shape=len(X), fill_value=self.classes_[0],
-            dtype=self.classes_.dtype
+            shape=len(X),
+            fill_value=self.classes_[0],
+            dtype=self.classes_.dtype,
         )
 
-        # XXX fix
+        # pour chaque ligne (point) de X
+        #   calculer la distance entre le point et tous les autres points.
+        #   garder le point avec la plus petit distance
+        #   retourner sa classe
+
+        # le faire si possible avec le broadcasting
+
+        # la distance avec le broadcasting mais que pour le premier point
+        np.sqrt(np.sum((self.X_train_ - X[0]) ** 2, axis=1))
+
+        # On veut un tableau (n_test,n_train,p) le p est le même car on a le
+        # même nombre de features
+
+        # On a un tableau ou la distance au carré entre le point i et j est
+        # à la case (i,j)
+        # J'ai enlevé le sqrt puisque que pas utile pour les comparaisons
+        distances = np.sum(
+            (self.X_train_[None, :, :] - X[:, None, :]) ** 2, axis=2
+        )
+
+        # Maintenant on veut trouver pour chaque ligne le point
+        # avec la plus petite distance
+        closest_points = np.argmin(distances, axis=1)
+        # Ça retourne pour chaque ligne l'indice de la plus petite distance
+        y_pred = self.y_train_[
+            closest_points
+        ]  # et ça ça doit être les classes de ces points.
+
         return y_pred
 
     def score(self, X, y):
-        """Write docstring.
+        """Return the score.
+
+        Parameters
+        ----------
+        X : ndarray of shape (n_samples, n_features)
+            The input array.
+        y : ndarray of shape (n_samples,)
+            The output array.
+
+        Returns
+        -------
+        score : float,
+            the accuracy of the model between 0 and 1.
 
-        And describe parameters
         """
         X, y = check_X_y(X, y)
         y_pred = self.predict(X)
+        score = np.sum(y_pred == y) / len(y)
 
-        # XXX fix
-        return y_pred.sum()
+        return score