词云（WordCloud）制作

Michael阿明

发布于 2021-02-19 15:12:34

7750

发布于 2021-02-19 15:12:34

以《神雕侠侣》为例，我们制作词云，看看有哪些高频词汇。

1. 导入一些包

# -*- coding:utf-8 -*-
# @Python Version: 3.7
# @Time: 2020/11/27 19:32
# @Author: Michael Ming
# @Website: https://michael.blog.csdn.net/
# @File: word_cloud.py
# @Reference: 

import jieba
import numpy as np
from PIL import Image
from wordcloud import WordCloud

2. 文本处理

去掉一些空白的字符\n, \t, ' '等
jieba.cut分词
屏蔽停用词，以及长度小于2的词

stopwords = []
with open("stopwords.txt", 'r', encoding='utf-8') as f:
    for w in f:
        stopwords.append(w.replace('\n', ''))

with open('processed_txt.txt', 'w', encoding='utf-8') as processed:
    words_list = []
    with open("../shendiaoxialv.txt", 'r', encoding='utf-8') as f:
        for line in f:
            word = jieba.cut(line.replace('\t', '').replace('\n', '').replace(' ', ''))
            for w in word:
                if w not in stopwords and len(w) > 1:
                    words_list.append(w)
    processed.write(' '.join(words_list))

3. 制作词云

# 读取清理好了以后的词
wordtxt = ""
with open('processed_txt.txt', 'r', encoding='utf-8') as f:
    wordtxt = f.read()

# 设置背景图片，也可以没有
background_pic = np.array(Image.open('background.png'))
wc = WordCloud(
    background_color='white', # 背景色
    max_words=300, # 最多词语数量
    font_path='wb.ttf', # 字体
    min_font_size=12, # 最小字号
    max_font_size=52, # 最大字号
    width=1600, # 图片宽度
    height=1200, # 图片高度
    mask=background_pic # 背景形状
)

wc.generate(wordtxt)
wc.to_file("wc.jpg")

使用以下背景图片：