分词及词云图设计

１.jieba的基本使用

import jieba


s1 = '我喜欢广州小蛮腰'
s2 = "我喜欢上海东方明珠"
#jieba,cut()#默认精准模式
print(10*'-','全模式',10*'-')
r1 = jieba.cut(s1,cut_all=True)#全模式
print(s1)
for ele in r1:
    print(ele)

print(10*'-','精准模式',10*'-')
r2 = jieba.cut(s1,cut_all=False)
print(s1)
for ele in r2:
    print(ele)

print(10 * '-', '搜索引擎模式', 10 * '-')
r3 = jieba.cut_for_search(s1)#搜索引擎模式
for ele in r3:
    print(ele)

#词性标注
import jieba.posseg
print('---词性标注---')
r4 = jieba.posseg.cut(s1)
# flag 词性
# word 词语
for ele in r4:
    print(ele.word+ele.flag)

#词典加载
jieba.load_userdict('/home/chen/anaconda3/lib/python3.6/site-packages/jieba/dict.txt')
#更改词频
print('---更改词频---')
word_chg = '上海东方'
jieba.add_word(word_chg)
jieba.suggest_freq(word_chg,True)

#提取关键词
import jieba.analyse
tag = jieba.analyse.extract_tags(s2,3)
print(tag)

#返回词语的位置
print('---返回词语的位置---')
word_index = jieba.tokenize(s2)
for ele in word_index:
    print(ele)
print('---以搜索引擎的方式')
Word_index = jieba.tokenizes(s2,mode='search')
for ele in word_index:
    print(ele)
'''
a 形容词
c 连词
d 副词
e 叹词
f 方位词
i 成语
m 数量词
n 名词
nr 人名
ns 地名
nt 机构团体
nz 其他专有名词
p 介词
r 代词
t 时间
u 助词
v 动词
vn 动名词
w 标点
un 未知词语

２.文本挖掘实例

import jieba.analyse
import os

print(os.path.dirname(os.getcwd()))
file_name = '/home/chen/projects/词云/resource/斗罗大陆.txt'
path = os.path.join(os.path.dirname(os.getcwd()),file_name)

data = open(file_name).read()
#默认20个关键词
tag = jieba.analyse.extract_tags(data,50)#提取50个关键词
print(tag)

３.词云图设计

#收据收集-->分词-->过滤噪音词-->筛选高频词-->生成图

from wordcloud import WordCloud
import os
import jieba  #中文分词
import numpy as np
import PIL.Image as img #图片处理
import random


stopwords = {"然后":0,"这些":0,"那些":0,"如果":0}
cur_path = os.path.dirname(os.path.dirname(__file__)) + "/resource"

def chinese_jieba(text):
    wordlist_jieba=jieba.cut(text)
    text_jieba=" ".join(wordlist_jieba)
    return text_jieba

with open(os.path.join(cur_path,'test.txt')) as fp:
    text = fp.read()
    text = chinese_jieba(text)
    r_num = random.randrange(0,5)
    r_num = 4
    if r_num == 0:
        bg_img = 'zh.png'
    elif r_num == 1:
        bg_img = 'heart.png'
    elif r_num == 2:
        bg_img = 'broken_heart.png'
    elif r_num == 3:
        bg_img = 'tree.png'
    else:
        bg_img = 'love_love.png'
    #设置背景图
    mask_pic = np.array(img.open(os.path.join(cur_path,bg_img)))
    #默认生成方块词云,背景颜色为黑色
    # wcd = WordCloud().generate(text)
    #指定背景颜色,指定关键词个数,设置最大字体大小,设置字体,过滤词,背景图形
    wcd = WordCloud(background_color='white',
                    #max_words=100,
                    max_font_size=40,
                    font_path='JDFZONGYI.ttf',
                    stopwords=stopwords,
                    mask=mask_pic).generate(text)
    image=wcd.to_image()
    image.show()