如何根据 pandas 数据框中的频率创建词云

Question

我要画一个词云。 'tweets.csv' 是一个 Pandas 数据框，它有一个名为 'text' 的列。绘制的图表并不是基于最常见的词，艰难。单词大小如何与它们在数据框中的频率相关联？

text = df_final.text.values
wordcloud = WordCloud(
    #mask = logomask,
    max_words = 1000,
    width = 600,
    height = 400,
    #max_font_size = 1000,
    #min_font_size = 100,
    normalize_plurals = True,
    #scale = 5,
    #relative_scaling = 0,
    background_color = 'black',
    stopwords = STOPWORDS.union(stopwords)
).generate(str(text))
fig = plt.figure(
    figsize = (50,40),
    facecolor = 'k',
    edgecolor = 'k')
plt.imshow(wordcloud, interpolation = 'bilinear')
plt.axis('off')
plt.tight_layout(pad=0)
plt.show()

我的数据框如下所示：

0   RT @Pontifex_pt: Temos que descobrir as riquezezas ...
1   RT @Pontifex_pt: Todos estamos em viagem rumo ...
2   RT @Pontifex_pt: Unamos as forças, em todos ...
3   RT @GeneralMourao: #Segurançapública, preocupa ...
4   RT @FIFAcom: The Brasileirao U-17 final provided ...

Answer 1

设置示例 DataFrame：

• 另见 DataCamp: Generating WordClouds in Python
• 软件包文档位于 WordCloud for Python documentation, GitHib: wordcloud, and Anaconda: wordcloud

import pandas as pd

df = pd.DataFrame({'word': ['how', 'are', 'you', 'doing', 'this', 'afternoon'],
                   'count': [7, 10, 4, 1, 20, 100]}) 

        word  count
0        how      7
1        are     10
2        you      4
3      doing      1
4       this     20
5  afternoon    100

将 word 和 count 列转换为 dict

• WordCloud().generate_from_frequencies() 需要 dict
• 使用以下方法之一

# method 1: convert to dict 
data = dict(zip(df['word'].tolist(), df['count'].tolist()))

# method 2: convert to dict
data = df.set_index('word').to_dict()['count']

print(data)

[out]: {'how': 7, 'are': 10, 'you': 4, 'doing': 1, 'this': 20, 'afternoon': 100}                                                                          

词云：

• 使用.generate_from_frequencies
• generate_from_frequencies(frequencies, max_font_size=None)

from wordcloud import WordCloud

wc = WordCloud(width=800, height=400, max_words=200).generate_from_frequencies(data)

情节

import matplotlib.pyplot as plt

plt.figure(figsize=(10, 10))
plt.imshow(wc, interpolation='bilinear')
plt.axis('off')
plt.show()

使用图像遮罩：

twitter_mask = np.array(Image.open('twitter.png'))
wc = WordCloud(background_color='white', width=800, height=400, max_words=200, mask=twitter_mask).generate_from_frequencies(data_nyt)

plt.figure(figsize=(10, 10))
plt.imshow(wc, interpolation='bilinear')
plt.axis("off")
plt.figure()
plt.imshow(twitter_mask, cmap=plt.cm.gray, interpolation='bilinear')
plt.axis("off")
plt.show()