import jieba
from jieba.analyse import extract_tags
from scipy.misc import imread
from wordcloud import WordCloud

import matplotlib.pyplot as plt

def generater(**kwargs):
    content_name = kwargs['content_name']
    content_path = './{}.txt'.format(content_name)
    top_k = kwargs['top_k']
    bg_name = kwargs['bg_name']
    color = kwargs['color']
    font_type = kwargs['font_type']

    with open(content_path, mode='r', encoding='utf-8') as f:
        content = f.read()
    '''
    根據TF/IDF提取topK個關鍵詞
    '''
    tags = extract_tags(sentence=content, topK=top_k)
    '''
    得到關鍵詞的詞頻
    '''
    # 全模式
    words = [word for word in jieba.cut(content, cut_all=True)]
    words_freq = {}
    for tag in tags:
        freq = words.count(tag)
        words_freq[tag] = freq
    '''
    設定背景
    scipy.misc imread()：返回的是 numpy.ndarray 也即 numpy 下的多維陣列物件
    '''
    bg_path = './{}.png'.format(bg_name)
    bg_img = imread(bg_path)
    font_path = './{}.ttf'.format(font_type)
    word_cloud = WordCloud(font_path=font_path, # 設定字型
                           background_color=color, # 背景顏色
                           max_words=top_k, # 詞雲顯示的最多詞數
                           max_font_size=100, # 字型最大
                           mask=bg_img, # 背景圖
                           )
    word_cloud.generate_from_frequencies(words_freq)

    plt.imshow(word_cloud)
    plt.axis('off') # 不顯示座標軸
    plt.show()

    # 儲存圖片
    word_cloud_img = './{}_word_cloud.jpg'.format(content_name)
    word_cloud.to_file(word_cloud_img)

if __name__ == '__main__':
    generater(content_name='習近平：在慶祝中國共產黨成立95週年大會上的講話',
              top_k=66,
              bg_name='china', # 預設png
              color='black',
              font_type='wryh')



 

windows 安裝 wordcloud

下載地址

pip install  wordcloud-1.3.1-cp36-cp36m-win_amd64.whl