综合练习:词频统计

时间:2023-02-13 12:59:02
text= open("aaa.txt","r")
ls=text.read()
text.close()
print(ls)
str = ''',.'"!?:'''
enclude={'for','you','a','and','the',}
for i in str:
    ls = ls.replace(i," ")
    # print(ls)
ls=ls.lower().split()
print(ls)
lsDict={}
lsset = set(ls)-enclude
for i in lsset:
    lsDict[i]=ls.count(i)
for w in lsDict:
    print(w,lsDict[w])

# for i in ls:
#     lsDict[i] = lsDict.get(i,0)+1
# for w in lsDict:
#     print(w,lsDict[w])

dictlist = list(lsDict.items())
dictlist.sort(key=lambda x:x[1],reverse=True)
for i in dictlist:
    print(i)

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

综合练习:词频统计