绘制新浪财经上与“贵州茅台”相关的新闻的词云图。

1 需求

绘制新浪财经上与“贵州茅台”相关的新闻的词云图。

2 代码实现

import requests
import re
import jieba
import numpy as np
from imageio import imread
from PIL import Image
from wordcloud import WordCloud, ImageColorGenerator


headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101 Firefox/91.0'}
url = 'https://search.sina.com.cn/news?q=贵州茅台&c=news&sort=time'
res = requests.get(url=url, headers=headers).text
try:
    res = res.encode('ISO-8859-1').decode('utf-8')
except:
    try:
        res = res.encode('ISO-8859-1').decode('gbk')
    except:
        res = res
p_title = '<h2><a href.*?target="_blank">(.*?)</a>'
title = re.findall(p_title, res)

for index in range(len(title)):
    title[index] = re.sub('<.*?>', '', title[index]).strip()
content = ' '.join(title)
contents = jieba.cut(content)
reports = []
for content in contents:
    if len(content) >= 2:
        reports.append(content)

# 按特定形状和特定颜色绘制词云图
# 获取词云图的形状蒙版
background_pic = '酒.jpg'
images = Image.open(background_pic)
maskImages = np.array(images)

# 按照形状蒙版绘制词云图
content = ' '.join(reports)
wc = WordCloud(
    font_path='simhei.ttf',
    background_color='white',
    width=1000,
    height=600,
    mask=maskImages
).generate(content)

# 修改词云图的颜色
back_color = imread(background_pic, format=None)
image_colors = ImageColorGenerator(back_color)
wc.recolor(color_func=image_colors)

wc.to_file('词云图+自定义+颜色.png')