Python using micro-channel buddy information acquired itchat ~

Recently discovered a fun package itchat , the interface is implemented by calling the micro-letter web version of sending and receiving messages, get some friends function information, you can be the venue itchat Projects view details.

aims:

• Get a buddy list
• Statistical sex and urban distribution

• According to a friend signature generation word cloud

  Friends get information

• Earlier it said, itchat actually call micro-letter web version of the interface, so when landing landing two-dimensional code will pop up, then itchat.get_friends get information like friends, is quite simple.

# -*- coding: utf-8 -*-
import itchat
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import warnings
import jieba
import re
from scipy.misc import imread  
from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator
itchat.login()
#登陆网页版微信，需要手机扫码确认
warnings.filterwarnings("ignore")

• Friends get information when the need to note that, because the first one is his buddy list, so start from second place to save.

friends = itchat.get_friends(update=True)
for counter,content in enumerate(friends[1:]):
    if counter == 0:
        df=pd.DataFrame(content)
        df.columns=content.keys()
    else:
        df.loc[counter]=content.values()
print '获取到%d位好友信息'%counter
df.columns

• Field information is as follows:

Index([u'UserName', u'City', u'DisplayName', u'UniFriend', u'MemberList',
       u'PYQuanPin', u'RemarkPYInitial', u'Sex', u'AppAccountFlag',
       u'VerifyFlag', u'Province', u'KeyWord', u'RemarkName', u'PYInitial',
       u'IsOwner', u'ChatRoomId', u'HideInputBarFlag', u'EncryChatRoomId',
       u'AttrStatus', u'SnsFlag', u'MemberCount', u'OwnerUin', u'Alias',
       u'Signature', u'ContactFlag', u'NickName', u'RemarkPYQuanPin',
       u'HeadImgUrl', u'Uin', u'StarFriend', u'Statues'],
      dtype='object')

Friends city distribution

• Because the micro-channel is a two-position information, provinces and cities, Beijing and Shanghai for this municipality, City save field information area, this part of the deal made under the name of the city changed accordingly.

df['City'][df['Province']==u'北京']= u'北京'
df['City'][df['Province']==u'上海']= u'上海'
df['City'][df['Province']==u'重庆']= u'重庆'
#微信对于直辖市将city字段填写为区
plt.figure(figsize = (15,12))
data_temp = df[df['City']!='']
#剔除城市未填写的记录
data_temp = data_temp.groupby(['City'])['UserName'].count().reset_index()
data_temp = data_temp.nlargest(15,'UserName')
data_temp.columns = ['City','Count']
sns.barplot(data=data_temp ,y='City',x='Count')
for y,x in enumerate(data_temp['Count']):
    plt.text(x,y,x,fontsize = 20)
plt.title(u'好友城市分布',fontsize =25)
plt.show()

• Just take the top 15, the effect is as follows:
  

  Gender distribution

• Sex fields have 0,1,2 three values, representing the gender is unknown, male, female , corresponding replace just fine.

df['Sex'] = df['Sex'].replace({0:u'性别不明',1:u'男',2:u'女'})
plt.figure(figsize = (15,8))
plt.subplot(1,2,1)
data_temp = df.groupby(['Sex'])['UserName'].count().reset_index()
data_temp = data_temp.sort_values('UserName')
data_temp.columns = ['Sex','Count']
sns.barplot(data=data_temp ,x='Sex',y='Count')
for x,y in enumerate(data_temp['Count']):
    plt.text(x-0.05,y,y,fontsize = 20)
plt.subplot(1,2,2)
sex_list = [u'男',u'女',u'性别不明']
countlist = [292,227,25]
explode = (0.05,0,0)
plt.pie(countlist,labels = sex_list,explode =explode,startangle = 90,autopct = '%3.1f%%')
plt.axis('equal')
plt.show()

• Results are as follows:
  

  According to a friend signature word cloud draw

• Signature stored in the signature are friends, I began tried again and found the signature emoji expression accounted for a big proportion, followed by regular expression handling the case, leaving only a part of Chinese characters.

back_color = imread('tencent.jpg')  # 解析该图片
wc = WordCloud(background_color='white',  # 背景颜色
               max_words=1000,  # 最大词数
               mask=back_color,  # 以该参数值作图绘制词云，这个参数不为空时，width和height会被忽略
               max_font_size=100,  # 显示字体的最大值
               font_path="/Users/tangwenpan/Downloads/simhei.ttf",  # 解决显示口字型乱码问题
               random_state=42,  # 为每个词返回一个PIL颜色
               )

text=''
xx= u"[\u4e00-\u9fa5]" #保留汉字
for x in df['Signature']:
    pattern = re.compile(xx)  
    text_temp =  pattern.findall(x) 
    for xxx in text_temp:
        text = text +xxx

def word_cloud(texts):
    words_list = []
    word_generator = jieba.cut(texts, cut_all=False)  # 返回的是一个迭代器
    for word in word_generator:
        if len(word) > 1:  #去掉单字
            words_list.append(word)
    return ' '.join(words_list)  


text = word_cloud(text)

wc.generate(text)
# 基于彩色图像生成相应彩色
image_colors = ImageColorGenerator(back_color)
plt.figure(figsize = (15,15))
plt.axis('off')
# 绘制词云
plt.imshow(wc.recolor(color_func=image_colors))
plt.axis('off')
# 保存图片
wc.to_file('comment.png')
print 'comment.png has bee saved!'

• Use background picture
  )
• Results are as follows:
  
  ----
  Peace ~