第3章 k近邻法(KNearestNeighbors)代码实现

=============================== 【回到目录】===============================

第3章 k近邻法(KNearestNeighbors)代码实现

def L(x, y, p=2):
    # x1 = [1, 1], x2 = [5,1]
    if len(x) == len(y) and len(x) > 1:
        sum = 0
        for i in range(len(x)):
            sum += math.pow(abs(x[i] - y[i]), p)
        return math.pow(sum, 1/p)
    else:
        return 0

# 课本例3.1
x1 = [1, 1]
x2 = [5, 1]
x3 = [4, 4]

# x1, x2
for i in range(1, 5):
    r = { '1-{}'.format(c):L(x1, c, p=i) for c in [x2, x3]}
    # print(r)
    # for a, b in zip(r.values(), r.keys()):
    #     print(a, b)
    # print(min(zip(r.values(), r.keys())))
    print(min(zip(r.keys(), r.values())))
    # print("============================")

outcome：

('1-[4, 4]', 6.0)
('1-[4, 4]', 4.242640687119285)
('1-[4, 4]', 3.7797631496846193)
('1-[4, 4]', 3.5676213450081633)

KNN实现分类

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from collections import Counter

# data
iris = load_iris()
df = pd.DataFrame(iris.data, columns=iris.feature_names)
df['label'] = iris.target
df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']
# data = np.array(df.iloc[:100, [0, 1, -1]])

plt.figure(figsize=(15, 8))
plt.subplot(121)
plt.scatter(df[:50]['sepal length'], df[:50]['sepal width'], label='0')
plt.scatter(df[50:100]['sepal length'], df[50:100]['sepal width'], label='1')
plt.xlabel('sepal length')
plt.ylabel('sepal width')
plt.legend()
plt.title('original data')

data = np.array(df.iloc[:100, [0, 1, -1]])
X, y = data[:,:-1], data[:,-1]
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)



class KNN:
    def __init__(self, X_train, y_train, n_neighbors=3, p=2):
        """
        parameter: n_neighbors 临近点个数
        parameter: p 距离度量
        """
        self.n = n_neighbors
        self.p = p
        self.X_train = X_train
        self.y_train = y_train

    def predict(self, X):
        # 取出n个点
        knn_list = []
        for i in range(self.n):
            dist = np.linalg.norm(X - self.X_train[i], ord=self.p)
            knn_list.append((dist, self.y_train[i]))

        for i in range(self.n, len(self.X_train)):
            max_index = knn_list.index(max(knn_list, key=lambda x: x[0])) #这里的knn_list相当于是一个二维数组，所以lambda中取x[0]
            dist = np.linalg.norm(X - self.X_train[i], ord=self.p)
            if knn_list[max_index][0] > dist:
                knn_list[max_index] = (dist, self.y_train[i]) #注意这里的括号


        # 统计
        knn = [k[-1] for k in knn_list]
        count_pairs = Counter(knn) #count_pairs为键值对的形式
        max_count = sorted(count_pairs.items(), key=lambda x: x[1])[-1][0] #这里的lambda变量取得是x[1]
        return max_count



    def score(self, X_test, y_test):
        right_count = 0
        n = 10
        for X, y in zip(X_test, y_test):
            label = self.predict(X)
            if label == y:
                right_count += 1
        return right_count / len(X_test)


clf = KNN(X_train, y_train)
print('Test_score: {}'.format(clf.score(X_test, y_test)))
test_point = [6.0, 3.0]
print('Test Point: {}'.format(clf.predict(test_point)))

plt.subplot(122)
plt.scatter(df[:50]['sepal length'], df[:50]['sepal width'], label='0')
plt.scatter(df[50:100]['sepal length'], df[50:100]['sepal width'], label='1')
plt.plot(test_point[0], test_point[1], 'bo', label='test_point')
plt.xlabel('sepal length')
plt.ylabel('sepal width')
plt.legend()
plt.title('classification by KNN(n=3, p=2)')
plt.subplots_adjust(top=0.92, bottom=0.08, left=0.10, right=0.95, hspace=0.25,
                        wspace=0.35) #调整子图间距
plt.savefig("demo.jpg")
plt.show()

output:

Test_score: 1.0
Test Point: 1.0

kd-Tree

建立kd-Tree来进行分类

import numpy as np
from math import sqrt
import pandas as pd
from sklearn.datasets import load_iris
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split

iris = load_iris()
df = pd.DataFrame(iris.data, columns=iris.feature_names)
df['label'] = iris.target
df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']


data = np.array(df.iloc[:100, [0, 1, -1]])
train, test = train_test_split(data, test_size=0.4)
x0 = np.array([x0 for i, x0 in enumerate(train) if train[i][-1] == 0])
x1 = np.array([x1 for i, x1 in enumerate(train) if train[i][-1] == 1])


def show_train():
    plt.scatter(x0[:, 0], x0[:, 1], c='pink', label='[0]')
    plt.scatter(x1[:, 0], x1[:, 1], c='orange', label='[1]')
    plt.xlabel('sepal length')
    plt.ylabel('sepal width')


class Node:
    def __init__(self, data, depth=0, lchild=None, rchild=None):
        self.data = data
        self.depth = depth
        self.lchild = lchild
        self.rchild = rchild


class KdTree:
    def __init__(self):
        self.KdTree = None
        self.n = 0
        self.nearest = None

    def create(self, dataSet, depth=0):
        if len(dataSet) > 0: #注意这里不能写 if dataSet之类
            m, n = np.shape(dataSet)
            self.n = n - 1
            axis = depth % self.n
            mid = int(m / 2)
            dataSetcopy = sorted(dataSet, key=lambda x: x[axis])
            node = Node(dataSetcopy[mid], depth)
            if depth == 0:
                self.KdTree = node
            node.lchild = self.create(dataSetcopy[:mid], depth+1)
            node.rchild = self.create(dataSetcopy[mid+1:], depth+1)
            return node
        return None

    def preOrder(self, node):
        if node is not None:
            print(node.depth, node.data)
            self.preOrder(node.lchild)
            self.preOrder(node.rchild)

    def search(self, x, count=1):
        nearest = []
        for i in range(count):
            nearest.append([-1, None])
        self.nearest = np.array(nearest)

        def recurve(node):
            if node is not None:
                axis = node.depth % self.n
                daxis = x[axis] - node.data[axis]
                if daxis < 0:
                    recurve(node.lchild)
                else:
                    recurve(node.rchild)

                dist = sqrt(sum((p1 - p2) ** 2 for p1, p2 in zip(x, node.data)))
                for i, d in enumerate(self.nearest):
                    if d[0] < 0 or dist < d[0]:
                        self.nearest = np.insert(self.nearest, i, [dist, node], axis=0)
                        self.nearest = self.nearest[:-1]
                        break #这里一定要break

                n = list(self.nearest[:, 0]).count(-1) #self.nearest为np.array类型
                if self.nearest[-n-1, 0] > abs(daxis):
                    if daxis < 0:
                        recurve(node.rchild)
                    else:
                        recurve(node.lchild)

        recurve(self.KdTree)

        knn = self.nearest[:, 1]
        belong = []
        for i in knn:
            belong.append(i.data[-1])
        b = max(set(belong), key=belong.count) #注意这个用法

        return self.nearest, b


kdt = KdTree()
kdt.create(train)
kdt.preOrder(kdt.KdTree)

score = 0
for x in test:
    # input('press Enter to show next:')
    show_train()
    plt.scatter(x[0], x[1], c='red', marker='x', label='test point')  # 测试点
    near, belong = kdt.search(x[:-1], 5)  # 设置临近点的个数
    if belong == x[-1]:
        score += 1
    print("test:")
    print(x, "predict:", belong)
    print("nearest:")
    for n in near:
        print(n[1].data, "dist:", n[0])
        plt.scatter(n[1].data[0], n[1].data[1], c='green', marker='+')  # k个最近邻点
    plt.legend()
    plt.show()

score /= len(test)
print("score:", score)

output：

test:
[5.5 4.2 0. ] predict: 0.0
nearest:
[5.4 3.9 0. ] dist: 0.31622776601683805
[5.7 3.8 0. ] dist: 0.4472135954999583
[5.4 3.7 0. ] dist: 0.5099019513592784
[5.3 3.7 0. ] dist: 0.5385164807134505
[5.1 3.8 0. ] dist: 0.5656854249492386

kd-Tree建立以及分类的第二种方法，这种感觉写的有些不简练，推荐第一种

# kd-tree每个结点中主要包含的数据结构如下
class KdNode(object):
    def __init__(self, dom_elt, split, left, right):
        self.dom_elt = dom_elt  # k维向量节点(k维空间中的一个样本点)
        self.split = split  # 整数（进行分割维度的序号）
        self.left = left  # 该结点分割超平面左子空间构成的kd-tree
        self.right = right  # 该结点分割超平面右子空间构成的kd-tree


class KdTree(object):
    def __init__(self, data):
        k = len(data[0])  # 数据维度

        def CreateNode(split, data_set):  # 按第split维划分数据集exset创建KdNode
            if not data_set:  # 数据集为空
                return None #注意这里的递归出口
            # key参数的值为一个函数，此函数只有一个参数且返回一个值用来进行比较
            # operator模块提供的itemgetter函数用于获取对象的哪些维的数据，参数为需要获取的数据在对象中的序号
            # data_set.sort(key=itemgetter(split)) # 按要进行分割的那一维数据排序
            data_set.sort(key=lambda x: x[split])
            split_pos = len(data_set) // 2  # //为Python中的整数除法
            median = data_set[split_pos]  # 中位数分割点
            split_next = (split + 1) % k  # cycle coordinates


            # 递归的创建kd树
            return KdNode(median, split,
                          CreateNode(split_next, data_set[:split_pos]),  # 创建左子树
                          CreateNode(split_next, data_set[split_pos + 1:]))  # 创建右子树

        self.root = CreateNode(0, data)  # 从第0维分量开始构建kd树,返回根节点




# KDTree的前序遍历
def preorder(root):
    print(root.dom_elt)
    if root.left:  # 节点不为空
        preorder(root.left)
    if root.right:
        preorder(root.right)

        # 对构建好的kd树进行搜索，寻找与目标点最近的样本点：

data = [[2,3],[5,4],[9,6],[4,7],[8,1],[7,2]]
kd = KdTree(data)
preorder(kd.root)


from collections import namedtuple
from math import sqrt
from collections import namedtuple

# 定义一个namedtuple,分别存放最近坐标点、最近距离和访问过的节点数
result = namedtuple("Result_tuple", "nearest_point  nearest_dist  nodes_visited")


def find_nearest(tree, point):
    k = len(point)  # 数据维度

    def travel(kd_node, target, max_dist):
        if kd_node is None:
            return result([0] * k, float("inf"), 0)  # python中用float("inf")和float("-inf")表示正负无穷

        nodes_visited = 1 #注意nodes_visited初始化的位置

        s = kd_node.split  # 进行分割的维度
        pivot = kd_node.dom_elt  # 进行分割的“轴”

        if target[s] <= pivot[s]:  # 如果目标点第s维小于分割轴的对应值(目标离左子树更近)
            nearer_node = kd_node.left  # 下一个访问节点为左子树根节点
            further_node = kd_node.right  # 同时记录下右子树
        else:  # 目标离右子树更近
            nearer_node = kd_node.right  # 下一个访问节点为右子树根节点
            further_node = kd_node.left

        temp1 = travel(nearer_node, target, max_dist)  # 进行遍历找到包含目标点的区域

        nearest = temp1.nearest_point  # 以此叶结点作为“当前最近点”
        dist = temp1.nearest_dist  # 更新最近距离

        nodes_visited += temp1.nodes_visited

        if dist < max_dist:
            max_dist = dist  # 最近点将在以目标点为球心，max_dist为半径的超球体内

        temp_dist = abs(pivot[s] - target[s])  # 第s维上目标点与分割超平面的距离
        if max_dist < temp_dist:  # 判断超球体是否与超平面相交
            return result(nearest, dist, nodes_visited)  # 不相交则可以直接返回，不用继续判断

        # ----------------------------------------------------------------------
        # 计算目标点与分割点的欧氏距离
        temp_dist = sqrt(sum((p1 - p2) ** 2 for p1, p2 in zip(pivot, target)))

        if temp_dist < dist:  # 如果“更近”
            nearest = pivot  # 更新最近点
            dist = temp_dist  # 更新最近距离
            max_dist = dist  # 更新超球体半径

        # 检查另一个子结点对应的区域是否有更近的点
        temp2 = travel(further_node, target, max_dist)

        nodes_visited += temp2.nodes_visited
        if temp2.nearest_dist < dist:  # 如果另一个子结点内存在更近距离
            nearest = temp2.nearest_point  # 更新最近点
            dist = temp2.nearest_dist  # 更新最近距离

        return result(nearest, dist, nodes_visited)

    return travel(tree.root, point, float("inf"))  # 从根节点开始递归

# ret = find_nearest(kd, [3,4.5])
# print(ret)


from time import clock
from random import random


# 产生一个k维随机向量，每维分量值在0~1之间
def random_point(k):
    return [random() for _ in range(k)]


# 产生n个k维随机向量
def random_points(k, n):
    return [random_point(k) for _ in range(n)]

N = 400000
t0 = clock()
kd2 = KdTree(random_points(3, N))  # 构建包含四十万个3维空间样本点的kd树
ret2 = find_nearest(kd2, [0.1, 0.5, 0.8])  # 四十万个样本点中寻找离目标最近的点
t1 = clock()
print("time: ", t1 - t0, "s")
print(ret2)

output：

Test_score: 0.95
Test Point: 1.0
[7, 2]
[5, 4]
[2, 3]
[4, 7]
[9, 6]
[8, 1]
time:  9.723352499999999 s
Result_tuple(nearest_point=[0.09855415845874649, 0.5008869302133826, 0.7968480873923645], nearest_dist=0.0035793373761071217, nodes_visited=48)

sklearn knn 用法

from sklearn import datasets, neighbors

iris = datasets.load_iris()
X = iris.data
y = iris.target
knn_clf = neighbors.KNeighborsClassifier()
knn_clf.fit(X, y)
predict = knn_clf.predict(X)

accuracy = (y == predict).astype(int).mean()
print(accuracy)

output：

0.9666666666666667