期末综合大作业:词频统计1

f=open('son.txt',mode='r',encoding='utf-8')
fText = f.read()

print(fText)

replaceList = ['?',',','.',';']
for c in replaceList:
sonText = sonText.replace(c,' ')

print(sonText)

print(sonText.split(' '))
bigList = sonText.split('')

sonSet = set(bigList)
print(sonSet)

bigDict = {}
for word in sonSet:
bigDict[word] = bigList.count(word)
print(bigDict)

wordCountList = list(bigDict.items())
print(wordCountList)
wordCountList.sort(key=lambda x:x[1],reverse=True)

for i in range(20):
print (wordCountList[i])
print (wordCountList[i])

bigCountFile = open('sonCount.txt',mode='a',encoding='utf-8')
for i in range(len(wordCountList)):
bigCountFile.write(str(wordCountList[i][1])+' '+wordCountList[i][0]+'\n')
bigCountFile.close()

猜你喜欢

转载自www.cnblogs.com/wssscs/p/9205809.html