2021-01-14

KNN算法（鸢尾花）**

话不多说了，上代码啦！

在这里插入代码片
#knn算法复习
from numpy import *
import operator
from os import listdir
#使用KNN算法分类
'''def file2matrix(filename):
    with open(filename)as file:
        array0Lines=file.readlines()
        number0fLines=len(array0Lines)
        returnMat=zeros((number0fLines,3))
        classLabelVector=[]
        index=0
        for line in array0Lines:
            line=line.split()
            returnMat[index,:]=line[0:3]
            classLabelVector.append(int(line[-1]))
            index+=1
    print(type(returnMat))
    return returnMat,classLabelVector
'''
#鸢尾花数据集分类
def file2matrix(filename):
    with open(filename) as file:
        array0Lines=file.readlines()
        number0fLines=len(array0Lines)
        returnMat=zeros((number0fLines,4))
        classLabelVector=[]
        index=0
        returnMat1=[]
        for line in array0Lines:
            line=line.strip()
            listFromLine = line.split(",")
            classLabelVector.append(listFromLine[-1])
            returnMat[index,:]=listFromLine[0:4]
            
            index+=1
        #print('classLabelVectorL:',classLabelVector)
    #print(returnMat)
    return returnMat,classLabelVector
def classify(inX,dataSet,labels,k):
    dataSetSize = dataSet.shape[0]    # 训练样本个数
    diffMat = tile(inX, (dataSetSize, 1)) - dataSet    # np.tile: 重复n次
    sqDiffMat = diffMat**2
    sqDistances = sqDiffMat.sum(axis=1)
    distances = sqDistances**0.5    # distance是inX与dataSet的欧氏距离
    sortedDistIndicies = distances.argsort()    # 返回排序从小到达的索引位置
    classCount = {}   # 字典存储k近邻不同label出现的次数
    for i in range(k):
        voteIlabel = labels[sortedDistIndicies[i]]
        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1    # 对应label加1，classCount中若无此key，则默认为0
    sortedClassCount = sorted(classCount.items(), key=operator.itemgetter(1), reverse=True)    # operator.itemgetter 获取对象的哪个维度的数据
    return sortedClassCount[0][0]    # 返回k近邻中所属类别最多的哪一类

def autoNorm(dataSet):
    minVals=dataSet.min(0)
    maxVals=dataSet.max(0)
    ranges=maxVals-minVals
    normDataSet=zeros(shape(dataSet))
    m=dataSet.shape[0]
    normDataSet=dataSet-tile(minVals,(m,1))
    normDataSet=normDataSet/tile(ranges,(m,1))
    # print('m:',m)
    # print('normDataSet:',normDataSet)
    # print('ranges:',ranges)
    # print('minVals:',minVals)
    # print('maxVals:',maxVals)
    #print('dataSet:',dataSet)
    return normDataSet,ranges,minVals
def classifyPerson():
    resultList=['不喜欢','一般般','超喜欢']
    percentTats=float(input("玩游戏所耗的时间百分比？:"))
    ffMiles=float(input('每周玩获得飞行常客里程数？:'))
    iceCream=float(input('每周消费冰淇凌的公升数？:'))
    datingDataMat,datingLabels=file2matrix(r"D:\learnpy\机器学习算法\datingTestSet2.txt")
    normMat,ranges,minVals=autoNorm(datingDataMat)
    inArr=array([ffMiles,percentTats,iceCream])
    #print('inArr-minVals:',(inArr-minVals)/ranges)
    classifierResult=classify((inArr-minVals)/ranges,normMat,datingLabels,3)
    #print('你喜欢人的程度：',resultList[classifierResult-1])
#classifyPerson()
def test():
    hoRatio=0.60#设置用来测试的样本占比列
    #dataSet,labels=file2matrix(r"D:\learnpy\机器学习算法\datingTestSet2.txt")
    dataSet,labels=file2matrix(r"D:\learnpy\机器学习算法\iris.data")
    normDataSet,ranges,minVals=autoNorm(dataSet)
    m=normDataSet.shape[0]
    numTest=int(m*hoRatio)
    errorCount=0.0
    for i in range(numTest):
        classifierResult=classify(dataSet[i,:],dataSet[numTest:m,:],labels[numTest:m],3)
        print('classifier result:%s,real answer:%s'%(classifierResult,labels[i]))
        if (classifierResult!=labels[i]):
            errorCount+=1.0
    print('the total error rate is:%f'%(errorCount/float(numTest)))
#手写数字识别
def image2vector(filename):
    returnVector=zeros((1,1024))
    with open(filename) as file:
        for i in range(32):
            line=file.readline()
            for j in range(32):
                returnVector[0,32*i+j]=int(line[j])
    return returnVector
 
#k-近邻算法
#手写数字识别的测试代码
def handWritingClassTest():
    labels=[]
    trainingFileList=listdir(r'D:\learnpy\机器学习算法\trainingDigits')     #获取训练集目录下的文件名
    m=len(trainingFileList)                        #计算训练样本个数
    dataSet=zeros((m,1024))                        #初始化数据集
    for i in range(m):
        fileName=trainingFileList[i]
        number=int(fileName.split('_')[0])         #number为每个样本的分类
        labels.append(number)
        dataSet[i,:]=image2vector(r'D:\learnpy\机器学习算法\trainingDigits\%s'%fileName)
    testFileList=listdir(r"D:\learnpy\机器学习算法\testDigits")
    mTest=len(testFileList)
    error=0.0
    for i in range(mTest):
        fileName=testFileList[i]
        number=int(fileName.split('_')[0])
        testVector=image2vector(r"D:\learnpy\机器学习算法\testDigits\%s"%fileName)
        resultNumber=classify(testVector,dataSet,labels,3)      #使用算法估计样本所属类别
        print("the classifier came back with: %d, the real answer is: %d" % (resultNumber, number))
        if number!=resultNumber:                                #算法结果与样本的实际分类做对比
            error+=1.0
    print('the total number of errors is:',error)
    print('the total error rate is:',error/mTest)
#handWritingClassTest()
   
if __name__=='__main__':
    #returnMat,classLabelVector=file2matrix(r"D:\learnpy\机器学习算法\datingTestSet2.txt")
    dataSet,labels=file2matrix(r"D:\learnpy\机器学习算法\iris.data")
    # autoNorm(returnMat)
    #handwritingClassTest()
    test() 

运行的时候注意文件路径，以免出错。

鸢尾花数据集合在uci有下载，如想要更多的数据集链接如下
uci数据集下载
运行结果@
仅供你们参考，如有不足，多多指点。