import pandas as pd
data = pd.read_table('val.txt',names=['category','theme','URL','content'])
#讀取資料，轉成DataFrame格式。因為前期資料已經是處理好的，所以可以這樣寫。這裡的read_table和read_csv可以互換的，一個意思。

data.dropna(inplace=True)  #丟掉有空資料的行
print(data.shape)  #發現沒有空的資料。。。
data_content = data['content'].tolist()  #把content一列變成list，等下用jieba分詞器拆解每一行
 

import jieba
data_content_list = []   #這一個list中套list，也就是二維陣列
for i in data_content:
    jieba_content = jieba.lcut(i)
    if len(jieba_content)>1 and jieba_content!='\r\n':   #排除空行之類的資料
        data_content_list.append( jieba_content )
data_content_df = pd.DataFrame({'data_content_list':data_content_list})

import pandas as pd
stopwords_data = pd.read_table("stopwords.txt",sep="\t",quoting=3,names=['stopword'],encoding='utf-8')
#缺少 quoting=3報錯
stopwords_data_list = stopwords_data['stopword'].tolist()
stopwords_list_letter = []
for i in range(ord('ａ'),ord('ｚ')):
    stopwords_list_letter.append(chr(i))
 
for i in range(ord('Ａ'),ord('Ｚ')):
    stopwords_list_letter.append(chr(i))
#原資料中包含這些英文字母，但是這不是簡單的英文字母，ascii碼值沒有在0-127之內
stopwords_data_list = stopwords_list_letter+stopwords_data_list  #重新組成新的停用詞list
data_content_and_stopwords_list = []  #去停用詞之後的content，最終結果也是一個list中套list
all_content_and_stopwords_list = []
#data_content_list 是需要去停用詞的list，是一個list中套list
for i in data_content_list:
    line_content = []
    for j in i:
        if j not in stopwords_data_list:
            line_content.append(j)
            all_content_and_stopwords_list.append(j)  #all_content_and_stopwords_list 內容是所有詞，做詞頻用
    data_content_and_stopwords_list.append(line_content)   #這是去停用詞之後的結果

#統計詞頻，轉為字典
data_temporary_all_content_and_stopwords_df = pd.DataFrame({'content':all_content_and_stopwords_list})
counts = data_temporary_all_content_and_stopwords_df['content'].value_counts()
df_counts = pd.DataFrame({'counts':counts})
dic = dict(zip(df_counts.index.tolist(),df_counts['counts'].tolist()))
print(dic)
#統計詞頻，轉為字典
#上邊寫這麼多很麻煩，可以直接用collection直接實現
# from collections import Counter
# dic = Counter(all_content_and_stopwords_list)
#如果想把dic轉成字典
#df = pd.DataFrame(dic.items(), columns=['key', 'value'])

畫詞雲圖：

from wordcloud import WordCloud
import matplotlib.pyplot as plt

wordcloud=WordCloud(font_path="./data/simhei.ttf",background_color="white",max_font_size=80)
wordcloud=wordcloud.fit_words(frequencies=dic)
#frequencies引數是接收的資料
plt.imshow(wordcloud)
#plt.axis('off')  #去掉橫縱座標
plt.show()