leventbass

from sklearn.datasets import load_iris
from KNearestNeighbors import KNearestNeighbors
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
import pandas as pd


dataset = load_iris()

X = dataset.data

leventbass

import numpy as np


class KNearestNeighbors():
    def __init__(self, X_train, y_train, n_neighbors=5, weights='uniform'):

        self.X_train = X_train
        self.y_train = y_train

        self.n_neighbors = n_neighbors

leventbass

kneighbors(X_test)

leventbass

data = np.hstack((X, y[:, np.newaxis]))
        
np.random.shuffle(data)

split_rate = 0.7

train, test = np.split(data, [int(split_rate*(data.shape[0]))])

X_train = train[:,:-1]
y_train = train[:, -1]

leventbass

from sklearn.datasets import make_classification
X, y = make_classification(n_samples = 1000, n_features=2, n_redundant=0, n_informative=2,
                             n_clusters_per_class=1, n_classes=3, random_state=21)

mu = np.mean(X, 0)
sigma = np.std(X, 0)

X = (X - mu ) / sigma

leventbass

def score(X_test, y_test):
    y_pred = predict(X_test)
        
    return float(sum(y_pred == y_test))/ float(len(y_test))

leventbass

def predict(X_test, weights='uniform'):
        
        class_num = 3
        
        if weights=='uniform':
            neighbors = kneighbors(X_test)
            y_pred = np.array([np.argmax(np.bincount(y_train[neighbor])) for neighbor in neighbors])
        
            return y_pred 
    

leventbass

def kneighbors(X_test, return_distance=False):
       
        n_neighbors = 5
        dist = []
        neigh_ind = []
        
        point_dist = [euclidian_distance(x_test, X_train) for x_test in X_test]

        for row in point_dist:
            enum_neigh = enumerate(row)

leventbass

def euclidian_distance(a, b):
    return np.sqrt(np.sum((a-b)**2, axis=1))

leventbass

import numpy as np
from sklearn.datasets import make_classification