[python基礎] python生成wordcloud並保存

1.核心包

#jieba、pandas用來處理數據，數據源以xls格式存儲的，這里用pandas進行處理
import jieba
from jieba import analyse
import pandas as pd
#scipy、wordcloud創建詞雲 from scipy.misc import imread
from wordcloud import WordCloud
from wordcloud import ImageColorGenerator
#matpoltlib展示、保存生成的詞雲圖
import matplotlib.pyplot as plt

2.過程

import jieba
from jieba import analyse import pandas as pd import sys reload(sys) sys.setdefaultencoding('utf-8') # 1.stopwords
def stop_words(): stop_dict = set() with open(u'./百度停用詞列表.txt', 'r')as f: words = f.readlines() for word in words: stop_dict.add(word.strip().decode('utf-8')) return stop_dict # 2.分詞並去停用詞 # save chinese only,remove english words,emoji
def remove_stopwords(stop_words): source_data = pd.read_excel('./11.xls') all_content = [] content = source_data[u'內容'] f = open('./weibo.txt', 'w') for line in content: cut_list = [c for c in jieba.cut(line)] ret_set = set(cut_list) - stop_words ret_list = list(ret_set) f.writelines([str(line) for line in ret_list]) f.writelines('\n') all_content.extend(ret_list) f.close() #3.統計詞頻
def get_frequency_words(file): with open(file, 'r') as f: texts = f.read()
　　　　　# 統計詞頻 top_words = analyse.textrank(texts, topK=400, withWeight=True) ret_words = {} for word in top_words: ret_words[word[0]] = word[1] return ret_words from scipy.misc import imread from wordcloud import WordCloud from wordcloud import ImageColorGenerator import matplotlib.pyplot as plt 
# 4.生成詞雲圖並保存 def generate_word_cloud(dict): color_mask = imread('./background.jpg') cloud = WordCloud( # 設置字體，不指定就會出現亂碼，文件名不支持中文
        font_path="./static/chinese.msyh.ttf", # font_path=path.join(d,'simsun.ttc'),
        # 設置背景色，默認為黑，可根據需要自定義為顏色
        background_color='white', # 詞雲形狀，
        mask=color_mask, # 允許最大詞匯
        max_words=400, # 最大號字體，如果不指定則為圖像高度
        max_font_size=150, # 畫布寬度和高度，如果設置了mask則不會生效
        # 詞語水平擺放的頻率，默認為0.9.即豎直擺放的頻率為0.1
        prefer_horizontal=0.8 ) cloud.generate_from_frequencies(frequencies=dict) cloud.to_file('word_cloud.jpg') # plt.imshow(cloud)
    # 不現實坐標軸
    plt.axis('off') # 繪制詞雲
    # plt.figure(dpi = 600)
    image_colors = ImageColorGenerator(color_mask)
 # 重新上色 plt.imshow(cloud.recolor(color_func=image_colors))
　　# 保存圖片  plt.savefig('./result2.png') # plt.show()

if __name__ == '__main__': stop_words = stop_words() remove_stopwords(stop_words=stop_words) words_frequency = get_frequency_words('./weibo.txt') generate_word_cloud(words_frequency) 

[注]：(1).wordcloud.generate_from_text(text=text)可以直接由文本生成詞雲，但必須是英文文本。

　　(2).wordcloud.generate_from_frequencies(frequencies=dict)由詞頻字典生成詞雲，詞頻越大則顯示該詞size越大
[結果]：