贝叶斯分类（涉及python3 的string的方法、文件读取操作方法、arff文件格式）

前段时间学习python3的基础知识，现在做一些数据挖掘方面的小练习。

今天要做的事贝叶斯分类器，数据来源为老师的weather.arff, 首先要读取这个文件，要用到string的方法http://write.blog.csdn.net/postlist和文件读取方法http://write.blog.csdn.net/postlist 和arff文件格式http://blog.sina.com.cn/s/blog_9d40b61301012xci.html，程序如下：

# coding =utf-8
import re
import sys

def readArff(fileName):
    arffFile = open(fileName,'r')
    data = []
    for line in arffFile.readlines():
        if not (line.startswith('@')):
            if not (line.startswith('%')):
                if line !='\n':
                    L=line.strip('\n')
                    k=L.split(',')
                    data.append(k)                    
                    print(k)
    print(data)
if __name__ =='__main__':
    fileName=r'C:\Users\Administrator\Desktop\exepirenment\classifill\data\weather.arff'
    readArff(fileName)

输出结果如下：

['sunny', '85', '85', 'FALSE', 'no']
['sunny', '80', '90', 'TRUE', 'no']
['overcast', '83', '86', 'FALSE', 'yes']
['rainy', '70', '96', 'FALSE', 'yes']
['rainy', '68', '80', 'FALSE', 'yes']
['rainy', '65', '70', 'TRUE', 'no']
['overcast', '64', '65', 'TRUE', 'yes']
['sunny', '72', '95', 'FALSE', 'no']
['sunny', '69', '70', 'FALSE', 'yes']
['rainy', '75', '80', 'FALSE', 'yes']
['sunny', '75', '70', 'TRUE', 'yes']
['overcast', '72', '90', 'TRUE', 'yes']
['overcast', '81', '75', 'FALSE', 'yes']
['rainy', '71', '91', 'TRUE', 'no']

读取完数据后，则开始写贝叶斯算法，贝叶斯原理不明白的请看我的新浪博客： http://blog.sina.com.cn/s/blog_bee847a20102v1wz.html。完整程序如下：

# coding =utf-8
import re
import sys
import bisect
data =[]  #全局变量
def readArff(fileName):
    arffFile = open(fileName,'r')
    global data
    for line in arffFile.readlines():
        if not (line.startswith('@')):
            if not (line.startswith('%')):
                if line !='\n':
                    L=line.strip('\n')
                    k=L.split(',')
                    data.append(k)
def bayesion(testData):
    class1=[]
    class2=[]
    global data
    for item in data:
        if item[len(item)-1] == 'yes':
            class1.append(item)
        else:
            class2.append(item)
    class1Probability = len(class1) /len(data)
    class2Probability = len(class2) /len(data)
    for i in range(len(testData)):
        count = 0
        for elem in class1:
            if testData[i]==elem[i]:
                count +=1       #统计个数
        class1Probability *= count/len(class1) #累计乘法 求总概率
        count = 0
        for elem in class2:
            if testData[i]==elem[i]:
                count +=1
        class2Probability *=count/len(class2)
    if class1Probability >class2Probability: #比较，进而分类
        print("The result is : Yes")
    else:
        print("The result if : No")
#数据预处理，将data数据分箱，data数据为 list[list1,list2...]类型
def dataPreprocessing1(data):
    breakpoint1 =[70,80]
    breakpoint2=[80,90]
    newValue1='LMH'
    for item in data:
        i = bisect.bisect(breakpoint1,int(item[1])) #int(),str()等类型需要转换
        item[1]=str(newValue1[i])
        j = bisect.bisect(breakpoint2,int(item[2]))
        item[2]=str(newValue1[j])
#数据预处理，分箱，针对单个list数据，data为list['..','..']类型
def dataPreprocessing2(data):
    breakpoint1 =[70,80]
    breakpoint2=[80,90]
    newValue1='LMH'
    i = bisect.bisect(breakpoint1,int(data[1]))
    data[1]=str(newValue1[i])
    j = bisect.bisect(breakpoint2,int(data[2]))
    data[2]=str(newValue1[j])


if __name__ =='__main__':
    fileName=r'C:\Users\Administrator\Desktop\exepirenment\classifill\data\weather.arff'
    readArff(fileName)
    dataPreprocessing1(data)

    testData =['overcast','72','80','TRUE']
    dataPreprocessing2(testData)

    bayesion(testData)

输出结果：

The result is : Yes