You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
56 lines
1.9 KiB
56 lines
1.9 KiB
"""
|
|
生成基于全部弹幕数据的词云图，并进行关键词提取和归一化处理
|
|
"""
|
|
|
|
import re
|
|
import pandas as pd
|
|
import numpy as np
|
|
import wordcloud
|
|
from matplotlib.image import imread
|
|
from jieba import analyse
|
|
from concurrent.futures import ThreadPoolExecutor
|
|
|
|
def blue_color_func(_, __, ___, ____, _random_state=None, **_kwargs):
|
|
"""定义蓝色调色板，用于词云图的颜色设置"""
|
|
return f"hsl(210, 100%, {np.random.randint(50, 90)}%)"
|
|
|
|
def normalize_hahaha(text):
|
|
"""归一化处理，将所有类似的“哈哈哈”统一为“哈哈哈”"""
|
|
return re.sub(r'哈{3,}', '哈哈哈', text)
|
|
|
|
def process_keywords(dm_list):
|
|
"""并行处理关键词提取"""
|
|
dm_string = ' '.join(dm_list)
|
|
with ThreadPoolExecutor() as executor:
|
|
keywords = list(executor.map(lambda kw: analyse.extract_tags(kw, topK=100, withWeight=False, allowPOS=()), [dm_string]))
|
|
return ' '.join(keywords[0])
|
|
|
|
def wordcloud_generation(danmu_data, stopwords, output_path):
|
|
"""生成词云图并保存"""
|
|
dm_list = danmu_data['danmu'].dropna().astype(str).tolist()
|
|
dm_list = [normalize_hahaha(text) for text in dm_list]
|
|
dmreal_string = process_keywords(dm_list)
|
|
|
|
img = imread("/output/OIP.jpg")
|
|
|
|
wc = wordcloud.WordCloud(
|
|
stopwords=stopwords,
|
|
width=1920,
|
|
height=1200,
|
|
background_color='white',
|
|
font_path='msyhl.ttc',
|
|
mask=img,
|
|
max_words=100,
|
|
color_func=blue_color_func,
|
|
).generate(dmreal_string)
|
|
wc.to_file(output_path)
|
|
|
|
def main():
|
|
"""加载数据并生成词云"""
|
|
dm = pd.read_excel('E:/Crawler/output/All_Danmu.xlsx', sheet_name='Sheet1')
|
|
stopwords = {'我', '你', '他', '这', '个', '是', '的', '了', '啊', '吗', '吧', '就', '都', '不', '也', '哈哈哈'}
|
|
wordcloud_generation(dm, stopwords, '/output/alldanmu_dwordcloud.png')
|
|
|
|
if __name__ == '__main__':
|
|
main()
|