画词云图（pyecharts和wordcloud）

本篇文章讲pyecharts和wordcloud两种方式画词云图，代码如下：
因为pyecharts的输入数据类型是列表中嵌套元组，类似下面这种的：

[('REX', 54), ('CASH+CARRY', 54), ('JUMBO', 824), ('SHOPPER', 147), ('BAG', 1951), ('VINTAGE', 1325), ('LEAF', 134), ('PEACE', 11), ('WOODEN', 507), ('BLOCK', 109)]

所以先把字符串describe_document变为字典，然后再list(字典.items())转为pyecharts需要的数据格式。

from pyecharts.charts import WordCloud

frequency = {}
for word in describe_document.split():
    if word not in frequency:
        frequency[word] = 1
    else:
        frequency[word] += 1
      
word_list = list(frequency.items())
word_list

mywordcloud = WordCloud()
mywordcloud.add('',word_list, shape='circle')

# 渲染图片
mywordcloud.render_notebook()

wordcloud库的generate函数可以直接对字符串文本进行分词

from wordcloud import WordCloud         
import matplotlib.pyplot as plt

# 数据采样
data = data.sample(20000,random_state = 22)

# 文本拼接
describe_document = " ".join(data['describe_cutted'])

fig = plt.figure(figsize=(20,10))

# 创建词云对象
wordcloud = WordCloud(background_color = "white",
              random_state = 30,
              scale = 2,
              collocations = False
              )

# 生成词云
wordcloud.generate(describe_document)
plt.imshow(wordcloud)
plt.axis("off")
plt.show()