kNN之手写数字识别

import numpy as np
# listdir()列出给定目录的文件名
from os import listdir
import operator


# inX-分类的输入向量,dataSet-输入的训练样本集,labels-标签向量,k-近邻数
def classify0(inX, dataSet, labels, k):
    dataSetSize = dataSet.shape[0]  # 得到训练集的行数,即样本个数
    # 以下三行距离计算计算
    # print("样本个数:",dataSetSize)
    diffMat = np.tile(inX, (dataSetSize, 1)) - dataSet  # tile():拉伸copy
    # 将输入的测试样本沿行方向扩充为(4,2),减去训练样本的坐标,得到各自的距离
    # print("变形:",diffMat)
    sqDiffMat = diffMat ** 2  # 欧式距离平方
    sqDistances = sqDiffMat.sum(axis=1)  # 欧式距离求和
    distances = sqDistances ** 0.5  # 开方
    sortedDistIndicies = distances.argsort()  # 按distances中的数值大小依次返回索引给y
    # print("索引顺序:",sortedDistIndicies)
    classCount = {}
    # 以下两行选择距离最小的k个点
    for i in range(k):
        voteIlabel = labels[sortedDistIndicies[i]]
        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1
        # 0:{labels[2]="B","B":1} 1:("B":2) 2:("A":1)
        # get():返回指定键的值，如果值不在字典中返回默认值(此处设为0)
    # 排序
    # print("classCount.items():",classCount.items())
    sortedClassCount = sorted(classCount.items(), key=operator.itemgetter(1), reverse=True)
    # operator.itemgetter(x):定义函数，获取对象的第x个域的值
    # key为函数，指定取待排序元素的哪一项进行排序
    # reverse = True 从大到小排序
    # print("sortedClassCount:",sortedClassCount)
    return sortedClassCount[0][0]


# 将图像转换为向量
def img2vector(filename):
    returnVect = np.zeros(1024)
    fr = open(filename)
    for i in range(32):
        lineStr = fr.readline()
        for j in range(32):
            returnVect[32 * i + j] = int(lineStr[j])  # 把逐行读取到的单行的每一项依次赋给
    return returnVect


def handwriteClassTest():
    hwLabels = []
    trainingFileList = listdir("trainingDigits")
    m = len(trainingFileList)  # 文件个数
    trainingMat = np.zeros((m, 1024))  # m个样本,每个样本1024个数据
    for i in range(m):
        # 以下三行从文件名解析分类数字
        fileNameStr = trainingFileList[i]  # 获取文件名
        fileStr = fileNameStr.split('.')[0]  # 文件名去后缀
        classNumStr = int(fileStr.split('_')[0])  # 获取文件数据所表示的值
        hwLabels.append(classNumStr)
        # 用样本值替换全0数组
        trainingMat[i, :] = img2vector('trainingDigits/%s' % fileNameStr)
    testFileList = listdir('testDigits')
    errorCount = 0.0
    mTest = len(testFileList)
    for i in range(mTest):
        fileNameStr = testFileList[i]
        fileStr = fileNameStr.split('.')[0]
        classNumStr = int(fileStr.split('_')[0])
        vectorUnderTest = img2vector('testDigits/%s'% fileNameStr)
        classifierResult = classify0(vectorUnderTest, trainingMat, hwLabels, 3)
        print("the classifier came back with: %d, the real answer is: %d" % (classifierResult, classNumStr))
        if (classifierResult != classNumStr):
            errorCount += 1.0
        print("
the total number of errors is: %d" % errorCount)
    print("
the total error rate is: %f" % (errorCount / float(mTest)))


if __name__ == "__main__":
    handwriteClassTest()