# coding:utf8

"""

DESC:NLTK词频统计

Author：伏草惟存

Prompt: code in Python3 env

"""

from nltk import *

from StopWords import readFile,seg_doc

#解决中文显示

import matplotlib

# （1）查看当前使用字体格式

# from matplotlib.font_manager import findfont, FontProperties

# print(findfont(FontProperties(family=FontProperties().get_family())))

# （2）在C:\Windows\Fonts查找中文字体SimHei.ttf，并将其复制到../mpl-data/font/ttf文件夹下面

# (3) 设置使用字体

matplotlib.rcParams['font.sans-serif'] = 'SimHei'

# 利用nltk进行词频特征统计

def nltk_wf_feature(word_list=None):

# ********统计词频方法1**************

fdist=FreqDist(word_list)

print(fdist.keys(),fdist.values())

print('='*3,'指定词语词频统计','='*3)

w='训练'

print(w,'出现频率：',fdist.freq(w)) # 给定样本的频率

print(w,'出现次数：',fdist[w]) # 出现次数

print('='*3,'频率分布表','='*3)

fdist.tabulate(10) # 频率分布表

# print('='*3,'可视化词频','='*3)

# fdist.plot(30) # 频率分布图

fdist.plot(30,cumulative=True) # 频率累计图

# print('='*3,'根据词语长度查找词语','='*3)

# wlist =[w for w in fdist if len(w)>2]

# print(wlist)

# ********统计词频方法2**************

from collections import Counter

Words = Counter(word_list)

print(Words.keys(),Words.values())

wlist =[w for w in Words if len(w)>2]

print(wlist)

return fdist

if __name__=='__main__':

# 1 读取文本

path= r'../Corpus/CSCMNews/体育/0.txt'

str_doc = readFile(path)

# print(str_doc)

# 2 词频特征统计

word_list =seg_doc(str_doc)

fdist = nltk_wf_feature(word_list)