Wordcloud 和列表列表

Question

我正在尝试使用以下数据创建词云：

my list= [[],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     ['EMF'],
     ['body'],
     [],
     [],
     [],
     [],
     ['water', 'juice'],
     ['What', 'are', 'u', 'doing'],
     [],
     [],
     [],
     [],
     [],
     [],
     [],
     ['EVENT'],
     ['christmas'],
     [],
     ['shalala'],
     ['happy'],
     []]

通常我会

import numpy as np
import pandas as pd
from PIL import Image
from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator

import matplotlib.pyplot as plt

corpus = " ".join(x for x in my_list)
df_wordcloud = WordCloud(background_color='white',max_font_size = 50).generate(corpus)

plt.imshow(df_wordcloud, interpolation='bilinear')
plt.axis("off")
plt.show()

但是这次我得到了错误：

TypeError: sequence item 0: expected str instance, list found

你知道如何从列表列表创建词云吗？

Answer 1

执行 " ".join(x for x in my_list) 实际上等效于 " ".join(my_list)，它试图将所有子列表连接在一起，导致您看到的错误。

相反，您想“扁平化”那些内部列表，首先将它们变成字符串：

corpus = " ".join(" ".join(x) for x in my_list)

甚至：

corpus = " ".join(map(" ".join, my_list))

Answer 2

让我们试试explode

s = pd.Series(l).explode().dropna()
Out[195]: 
19          EMF
20         body
25        water
25        juice
26         What
26          are
26            u
26        doing
34        EVENT
35    christmas
37      shalala
38        happy
dtype: object

Wordcloud 和列表列表

Wordcloud and list of list

python

pandas

matplotlib

word-cloud