《机器学习实战》系列博客主要是实现并理解书中的代码，相当于读书笔记了。毕竟实战不能光看书。动手就能遇到许多奇奇怪怪的问题。博文比较粗糙，需结合书本。博主边查边学，水平有限，有问题的地方评论区请多指教。书中的代码和数据，网上有很多请自行下载。

KNN算法的应用

2.2.1 从文本文件中解析数据

函数的输入为文件名字符串，输出为训练样本矩阵和类标签向量。

解析程序

def file2matrix(filename):fr = open(filename)numberOfLines = len(fr.readlines())         #get the number of lines in the filereturnMat = zeros((numberOfLines,3))        #prepare matrix to returnclassLabelVector = []                       #prepare labels return   fr = open(filename)index = 0for line in fr.readlines():line = line.strip()listFromLine = line.split('\t')returnMat[index,:] = listFromLine[0:3]  #选前三列的数据存到矩阵中classLabelVector.append(int(listFromLine[-1]))#最后一列转成整数后存到标签向量index += 1return returnMat,classLabelVector

导入数据成功，检查一下数据

>>> import kNN
>>> datingDataMat,datingLabels = kNN.file2matrix('datingTestSet2.txt')
>>> datingDataMat
array([[  4.09200000e+04,   8.32697600e+00,   9.53952000e-01],[  1.44880000e+04,   7.15346900e+00,   1.67390400e+00],[  2.60520000e+04,   1.44187100e+00,   8.05124000e-01],..., [  2.65750000e+04,   1.06501020e+01,   8.66627000e-01],[  4.81110000e+04,   9.13452800e+00,   7.28045000e-01],[  4.37570000e+04,   7.88260100e+00,   1.33244600e+00]])
>>> datingLabels[0:20]
[3, 2, 1, 1, 1, 1, 3, 3, 1, 3, 1, 1, 2, 1, 1, 1, 1, 1, 2, 3]

相关函数学习

• open 函数
  语法：open(name[, mode[, buffering]]) r:读操作；w：写操作；a:添加操作；b:二进制存取操作 如果缺省就是r

例如：在C:\Users\lzw\Desktop\python 目录下新建一个txt文件 test_open.txt

命令行输入，就可以打开这个文件

>>> f=open('C:\\Users\\lzw\\Desktop\\python\\test_open.txt','r+')  # "\" 是转义符，要将他再转义
>>> 

如果是不加路径，只有一个文件名：f = open (‘test_open.txt’) 则必须保证！！！ test_open.txt存储在我们的工作目录中

写文件操作

>>> f = open('test_open.txt', 'w')
>>> f.write('hello,')
>>> f.write('python\n')
>>> f.write('this is a test\n')
>>> f.write('lzw\n')
>>> f.close()
>>> 

打开 test_open.txt，可以看到原来空的txt 写入了内容

读文件操作

>>> f = open('test_open.txt')
>>> f.read(1)
'h'
>>> f.read(5)
'ello,'
>>> f.read()
'python\nthis is a test\nlzw\n'

readlines 逐行读取

>>> f = open('test_open.txt')
>>> f.readlines()
['hello,python\n', 'this is a test\n', 'lzw\n']
>>> 

• zeros 函数

>>> from numpy import*
>>> zeros(3)
array([ 0.,  0.,  0.])
>>> zeros((2,3))
array([[ 0.,  0.,  0.],[ 0.,  0.,  0.]])
>>> zeros([2,3])   #和上一种一样
array([[ 0.,  0.,  0.],[ 0.,  0.,  0.]])
>>> zeros(3,int16)  #默认的是float型
array([0, 0, 0], dtype=int16)
>>> a=array([[2,3],[3,4]])
>>> zeros_like(a)    #返回和输入大小相同，类型相同，用0填满的数组
array([[0, 0],[0, 0]])

• strip 函数 删除头尾字符串函数

>>> a = ' \n123\tabc\r'
>>> a.strip()  #删除头尾的字符串，默认为空白符(包括'\n', '\r',  '\t',  ' ')
'123\tabc'
>>> a.lstrip() #删除开头 
'123\tabc\r'
>>> a.rstrip() #删除结尾
' \n123\tabc'
>>> a.strip('12') 
' \n123\tabc\r'
>>> a = '12abc'
>>> a.strip('12')#删除字符串12
'abc'
>>> 

• split 函数拆分字符串。通过指定分隔符对字符串进行切片

>>> u = "www.doiido.com.cn"
>>> print u.split()  #使用默认分隔符
['www.doiido.com.cn']
>>> print u.split('.') #以"."为分隔符 
['www', 'doiido', 'com', 'cn']
>>> print u.split('.',1)  #分割一次
['www', 'doiido.com.cn']
>>> print u.split('.',2)  #分割两次
['www', 'doiido', 'com.cn']
>>> print u.split('.',2)[1]  #分割两次，并取序列为1的项
doiido
>>> u1,u2,u3 = u.split('.',2)#分割两次，并把分割后的三个部分保存到三个文件
>>> print u1
www
>>> print u2
doiido
>>> print u3
com.cn

2.2.2使用Matplotlib画散点图

kNN .py 程序里继续写，注意要import Matplotlib

散点图程序

datingDataMat,datingLabels = file2matrix('datingTestSet2.txt')
fig = plt.figure()
ax = fig.add_subplot(111)
ax.scatter(datingDataMat[:,1],datingDataMat[:,2],15.0*array(datingLabels),15.0*array(datingLabels))
plt.show()

相关函数学习

fig = plt.figure() #绘制一个figure对象

ax = fig.add_subplot(223) #将画布分割成2行2列，图像画在从左到右从上到下的第3块

scatter 画散点图

2.2.3归一化数值

newValue = (oldValue-min)/(max-min)

归一化程序

def autoNorm(dataSet):minVals = dataSet.min(0)  #每一列最小值maxVals = dataSet.max(0)ranges = maxVals - minValsnormDataSet = zeros(shape(dataSet)) #返回矩阵大小和数据矩阵一样用0填充m = dataSet.shape[0]  #矩阵行数normDataSet = dataSet - tile(minVals, (m,1))normDataSet = normDataSet/tile(ranges, (m,1))   #element wise dividereturn normDataSet, ranges, minVals

>>> normMat
array([[ 0.44832535,  0.39805139,  0.56233353],[ 0.15873259,  0.34195467,  0.98724416],[ 0.28542943,  0.06892523,  0.47449629],..., [ 0.29115949,  0.50910294,  0.51079493],[ 0.52711097,  0.43665451,  0.4290048 ],[ 0.47940793,  0.3768091 ,  0.78571804]])
>>> ranges
array([  9.12730000e+04,   2.09193490e+01,   1.69436100e+00])
>>> minVals
array([ 0.      ,  0.      ,  0.001156])
>>> 

2.2.4测试算法：用错误率来检测分类器的性能

分离器性能测试程序

def datingClassTest():hoRatio = 0.10      #选择10%数据作测试datingDataMat,datingLabels = file2matrix('datingTestSet2.txt') normMat, ranges, minVals = autoNorm(datingDataMat)m = normMat.shape[0]numTestVecs = int(m*hoRatio)errorCount = 0.0for i in range(numTestVecs):classifierResult = classify0(normMat[i,:],normMat[numTestVecs:m,:],datingLabels[numTestVecs:m],3)print "the classifier came back with: %d, the real answer is: %d" % (classifierResult, datingLabels[i])if (classifierResult != datingLabels[i]): errorCount += 1.0print "the total error rate is: %f" % (errorCount/float(numTestVecs))print errorCount

…

2.2.5构建完整可用的系统

预测程序

def classifyPerson():resultList = ['不喜欢','魅力一般的人','极具魅力的人']percentTats = float(raw_input("percentgage of time spent playing video game ?"))ffMile = float(raw_input("frequent flier miles earned per year ?"))iceCream = float(raw_input("liters of ice Cream consumed per year ?"))datingDataMat ,datingLabels = file2matrix('datingTestSet2.txt')normMat ,ranges ,minVals = autoNorm (datingDataMat)inArr = array([ffMile,percentTats,iceCream])classifierResult = classify0((inArr - minVals)/ranges, normMat,datingLabels,3)print "You will probably like this person: ",resultList[classifierResult-1]

>>> reload(kNN)
<module 'kNN' from 'kNN.py'>
>>> kNN.classifyPerson()
percentgage of time spent playing video game ? 10
frequent flier miles earned per year ?10000
liters of ice Cream consumed per year ?0.5
You will probably like this person:  魅力一般的人
>>> 

相关函数学习

raw_input() 将所有输入作为字符串看待，返回字符串类型