如何从 jupyter notebook 中的词云中删除自定义词

Question

wordcloud = WordCloud(background_color="white",width=1600,height=800).generate(' '.join(df1['text'].tolist()))
plt.figure(figsize=(20,10), facecolor='k')
plt.imshow(wordcloud)

The word cloud that gets shown has data that I don't think is relevant, such as user, need, anyone, issue, trying, and some other words. I have already removed stop words, but how can I remove a custom list of words from the word cloud? This is a word cloud from data I scraped from slack, and I want it to representative of questions people are asking in the channel that are more related to things like email, server, outlook, duo etc.

Answer 1

If you are using the wordcloud package, I think there isn't an option to remove user-provided custom words.

If there is a list of words you want to remove, I think you can remove them from df1 before fitting the word cloud.

Answer 2

You can do it like this:

• get the list of the default stopwords
• add your custom stopwords
• convert to a set
• use in WordClould

w = WordCloud()
stop_words = list(w.stopwords)
custom_stop_words = ['user', 'need']
stop_words = set(stop_words + custom_stop_words)

wordcloud = WordCloud(background_color="white",width=1600,height=800, stopwords=stop_words).generate(' '.join(df1['text'].tolist()))
plt.figure(figsize=(20,10), facecolor='k')
plt.imshow(wordcloud)