feat: add k-nearest neighbor machine learning algo

Python_codes/k-nearest neighbors/knn.py

@@ -0,0 +1,138 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+#
+# Copyright 2021 Jude Michael Teves
+#
+# This is my implementation of the k-nearest neighbors machine learning algorithm.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, see <http://www.gnu.org/licenses/>.
+
+import numpy as np
+import pandas as pd
+
+from collections import Counter
+
+class KNN():
+  def __init__(self, k:int=3, p:int=2):
+    '''
+    Parameters:
+    k - number of neighbors
+    p - order in Minkowski distance (1 - Manhattan, 2 - Euclidean)
+    '''
+    self.k = k
+    self.p = p
+    pass
+
+  def minkowski(self, X, point) -> list:
+    '''
+    Returns a list of distances of each elements in X from the reference point.
+    Vectorized implementation.
+    '''
+    X = np.array(X)
+    d = ((abs(X-point)**self.p).sum(axis=1))**(1/self.p)
+    distances = list(d)
+    return distances
+
+  def get_nearest_neighbors(self, X, point) -> list:
+    '''
+    Returns indices of nearest neighbors
+    '''
+    distances = self.minkowski(X, point)
+    indices = np.argsort(distances)[:self.k].tolist()
+    return indices
+
+  def fit(self, X, y):
+    '''
+    'Train' the model. In kNN's case, it just memorizes the values.
+    ---
+    Parameters:
+    X - input matrix
+    y - output vector
+    '''
+    self.X_train = X
+    self.y_train = y
+    pass
+
+  def predict(self, X) -> list:
+    '''
+    Return predictions
+    ---
+    Parameters:
+    X - input matrix
+    '''
+    pass
+
+  def score(self, X, y) -> float:
+    '''
+    R-squared for regression.
+    Accuracy for classification.
+    ---
+    Parameters:
+    X - input matrix
+    y - output vector
+    '''
+    pass
+
+
+class KNNRegressor(KNN):
+  @staticmethod
+  def r_squared(y_true:np.array, y_pred:np.array) -> float:
+    '''
+    Returns the score using R-squared metric.
+    '''
+    u = ((y_true - y_pred) ** 2).sum() # residual sum of squares
+    v = ((y_true - y_true.mean()) ** 2).sum() # total sum of squares
+    score = 1 - (u/v)
+    return score
+
+  def calculate(self, point):
+    '''
+    Returns the value of a new data point
+    '''
+    indices = self.get_nearest_neighbors(self.X_train, point)
+    labels = self.y_train[indices]
+    label = labels.sum()/len(labels)
+    return label
+
+  def predict(self, X) -> list:
+    y_pred = [self.calculate(point) for point in X]
+    return y_pred
+    pass
+
+  def score(self, X, y) -> float:
+    y_pred = self.predict(X)
+    return self.r_squared(y, y_pred)
+    pass
+
+
+class KNNClassifier(KNN):
+
+  def classify(self, point):
+    '''
+    Returns the classification of a new data point
+    '''
+    indices = self.get_nearest_neighbors(self.X_train, point)
+    labels = self.y_train[indices]
+    label = Counter(labels).most_common(1)[0][0]
+    return label
+
+  def predict(self, X) -> list:
+    y_pred = [self.classify(point) for point in X]
+    return y_pred
+    pass
+
+  def score(self, X, y) -> float:
+    y_pred = self.predict(X)
+    return sum(y_pred == y)/len(y) # accuracy
+    pass