中国近三年疫苗分布情况词云分析

数据来自github

from os import path
import jieba
import matplotlib.pyplot as plt
from wordcloud import WordCloud,STOPWORDS,ImageColorGenerator

text = ''
with open(r'C:\Users\Administrator\Desktop\疫苗名称.txt',encoding = 'utf8')as fin:
        for line in fin.readlines():
                line = line.strip('\n')
                text += ' '.join(jieba.cut(line))

backgroud_image = plt.imread(r'C:\Users\Administrator\Desktop\winter.jpg')#背景图

print('加载图片成功！')


wc = WordCloud(
    background_color='white',
    mask=backgroud_image,
    font_path='C:\Windows\Fonts\STZHONGS.TTF', 
    max_words=2000, 
    stopwords=STOPWORDS,
    max_font_size=150,
    random_state=30
)
wc.generate_from_text(text)
print('开始加载文本')
#改变字体颜色
img_colors = ImageColorGenerator(backgroud_image)
#字体颜色为背景图片的颜色
wc.recolor(color_func=img_colors)
# 显示词云图

plt.imshow(wc)
# 是否显示x轴、y轴下标
plt.axis('off')
plt.show()

d = path.dirname(__file__)

wc.to_file(path.join(d, "cj.jpg"))

print('生成词云成功!')

大多数疫苗都为乙型肝炎疫苗

 

看下公司分布

from os import path
import jieba
import matplotlib.pyplot as plt
from wordcloud import WordCloud,STOPWORDS,ImageColorGenerator

text = ''
with open(r'C:\Users\Administrator\Desktop\公司名称.txt',encoding = 'utf8')as fin:
        for line in fin.readlines():
                line = line.strip('\n')
                text += ' '.join(jieba.cut(line))

backgroud_image = plt.imread(r'C:\Users\Administrator\Desktop\winter.jpg')

print('加载图片成功！')

wc = WordCloud(
    background_color='white',
    mask=backgroud_image,
    font_path='C:\Windows\Fonts\STZHONGS.TTF', 
    max_words=2000, 
    stopwords=STOPWORDS,
    max_font_size=150,
    random_state=30
)
wc.generate_from_text(text)
print('开始加载文本')
#改变字体颜色
img_colors = ImageColorGenerator(backgroud_image)
#字体颜色为背景图片的颜色
wc.recolor(color_func=img_colors)
# 显示词云图

plt.imshow(wc)
# 是否显示x轴、y轴下标
plt.axis('off')
plt.show()
d = path.dirname(__file__)

wc.to_file(path.join(d, "cj.jpg"))

print('生成词云成功!')