python词频分析2

上一篇绘制了一个简单的词云图，下面在学习一些改进。

from wordcloud import WordCloud
import jieba
import imageio
with open('threekingdom.txt', 'r', encoding='utf-8') as f1:
    list1 = f1.read()
    list2 = jieba.lcut(list1)
    excludes = {"将军", "却说", "荆州", "二人", "不可", "不能", "如此", "如何", "东吴",
                "商议", "军马", "引兵", "次日", "主公", "左右", "大喜", "军士", "天下",
                "今日", "于是", "不敢", "魏兵", "都督", "陛下","玄德曰","孔明曰"
                }


    count={}#字典
    for word in list2:
      if len(word)<=1:
            continue
      else:
          count[word]=count.get(word,0)+1

    count["孔明"]=count["孔明"]+count["孔明曰"]
    count["玄德"] = count["玄德"] + count["玄德曰"]+count["刘备"]
    count["关公"] = count["关公"] + count["云长"]
    for word in excludes:
        del count[word]
    #print(count)
#转换成列表排序
    new_word_list=list(count.items())#元组
#print(new_word_list)
#进行排序
    new_word_list.sort(key=lambda x:x[1],reverse=True)
    text_list=[]
    for x in range(10):#显示前二十
     #print(new_word_list[x])
     role,count=new_word_list[x]#拆包
     print(role,count)
     for j in range(count):
        text_list.append(role)
     print(text_list)
     text=" ".join(text_list)
mask = imageio.imread("china.jpg")
wc = WordCloud(
        width=1000,
        height=800, 
        background_color='white',
        font_path="MSYH.TTC",
        mask=mask,
        collocations=False
    ).generate(text).to_file("三国词云.png")

上面的代码主要的改进是，把我不需要的词但是词频高的删除，之后再把重复的词从词云中删除，把同义词合并。注意wordcount的collocation的用法，还有这里的元组的转换，还有join函数，都是值得研究的点。