使用KNN识别MNIST手写数据集（手写，不使用KNeighborsClassifier）

import os

def load_data(path):
    check = [i for i in range(10)]
    final_data = []
    
    for i in range(10):
        final_data.append([])
        
    files = os.listdir(path)    #文件夹
    for file in files:
        data = open(path + "/" + file)
        str = ""      #将所有数据接在一起
        temp = []
        for line in data.readlines():
            str = str + line[:-1]   #去掉回车，一行接一行
        for i in str:
            temp.append(int(i))   #变成数字
        final_data[check.index(int(file[0]))].append(temp)   #根据标签放在列表相应的位置

    return final_data, len(files)

def knn_mnist(K,test_data):
    train_data, length = load_data('manifold/digits/trainingDigits')
    distance = []     #存储测试数据到所有训练数据的距离
    
    for i in range(len(train_data)):
        for j in range(len(train_data[i])):
            res = 0
            for k in range(len(test_data)):
                res += (test_data[k]-train_data[i][j][k]) ** 2   #欧氏距离
            distance.append([res ** 0.5, i])   #距离+训练集数据标签

    distance = sorted(distance, key=(lambda x: x[0]))  #按距离从小到大排序
    weight = []   #权重与序号
    sum_distance = 0.0
    for i in range(K):
        sum_distance += distance[i][0]   #计算前K个距离的和
    for i in range(K):
        weight.append([1 - distance[i][0] / sum_distance, distance[i][1]])  #权重+序号
        
    #将相同序号的加起来
    num = []   #统计有哪些序号
    for i in range(K):
        num.append(weight[i][1])
    num = list(set(num))   #去重
    
    final_res = []
    for i in range(len(num)):
        res = 0.0
        for j in range(len(weight)):
            if weight[j][1] == num[i]:   #前K个标签一样的样本权值加起来
                res += weight[j][0]
        final_res.append([res, num[i]])

    final_res = sorted(final_res, key=(lambda x: x[0]),reverse=True)  # 按照权重从大到小排序

    return final_res[0][1]   #最终返回最大权值对应的标签

def test():
    K = 5
    test_data, length = load_data('manifold/digits/testDigits')
    #测试
    for i in range(len(test_data)):
        for j in range(len(test_data[i])):
            print(knn_mnist(K, test_data[i][j]))

if __name__ == '__main__':
    test()