Lucien Descazeaud - numpy and sklearn assignments

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
 
 
@@ -40,28 +41,44 @@ def max_index(X):
     i = 0
     j = 0
 
-    # TODO
+    if not isinstance(X, np.ndarray):
+        raise ValueError("The input must be a numpy array")
+
+    d = X.ndim
+
+    if d != 2:
+        raise ValueError("The input must be a 2D array")
+
+    i_flat = np.argmax(X)
+
+    i, j = np.unravel_index(i_flat, X.shape)
 
     return i, j
 
 
 def wallis_product(n_terms):
-    """Implement the Wallis product to compute an approximation of pi.
-
-    See:
-    https://en.wikipedia.org/wiki/Wallis_product
+    """
+    Compute an approximation of pi using the Wallis product.
 
     Parameters
     ----------
     n_terms : int
-        Number of steps in the Wallis product. Note that `n_terms=0` will
-        consider the product to be `1`.
+        Number of terms in the Wallis product. If `n_terms=0`,
+        the function returns 1.
 
     Returns
     -------
     pi : float
-        The approximation of order `n_terms` of pi using the Wallis product.
+        Approximation of pi computed with `n_terms` 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
+
+    tab = np.arange(1, n_terms + 1)
+    n2 = tab * tab
+    n2x4 = 4 * n2
+    n2x4_1 = n2x4 - 1
+    final = n2x4 / n2x4_1
+    prod = np.prod(final)
+
+    return 2 * prod

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
@@ -29,7 +30,13 @@
 
 
 class OneNearestNeighbor(BaseEstimator, ClassifierMixin):
-    "OneNearestNeighbor classifier."
+    """
+    One-nearest-neighbor classifier.
+
+    This estimator stores the training data during `fit` and predicts
+    the label of a new sample as the label of the closest training
+    sample using the Euclidean distance.
+    """
 
     def __init__(self):  # noqa: D107
         pass
@@ -44,7 +51,9 @@ def fit(self, X, y):
         self.classes_ = np.unique(y)
         self.n_features_in_ = X.shape[1]
 
-        # XXX fix
+        self.X_ = X
+        self.y_ = y
+
         return self
 
     def predict(self, X):
@@ -55,11 +64,15 @@ def predict(self, X):
         check_is_fitted(self)
         X = check_array(X)
         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
+        for i in range(len(X)):
+            diff = self.X_ - X[i]
+            dist = np.linalg.norm(diff, axis=1)
+            idx = np.argmin(dist)
+            y_pred[i] = self.y_[idx]
+
         return y_pred
 
     def score(self, X, y):
@@ -71,4 +84,4 @@ def score(self, X, y):
         y_pred = self.predict(X)
 
         # XXX fix
-        return y_pred.sum()
+        return (y_pred == y).mean()