python调用bot登陆微信进行数据分析

python对微信进行数据分析

本文用到的包和工具如下：

wxpy

微信团队用python写的面向对象模拟机器人登陆网页微信服务器，最为重要的是bot类和itchat类；具体看：https://wxpy.readthedocs.io/zh/latest/

pyecharts

百度收购了一个图形可视化javascript前端项目Echarts，该项目又扩展为pyecharts，可以动态渲染出国家、省份、城市图等。具体看：http://pyecharts.herokuapp.com/

pillow

一个图像处理的包，用于拼接图片。具体看：https://pillow.readthedocs.io/en/latest/

将所有微信好友头像拼接

直接上python代码如下：

#导入需要的模块
from wxpy import *     #用于调用bot类来进行访问微信数据
import math
from PIL import Image  #用于拼接图片
import os              #用于来生成路径

# 创建头像存放文件夹
def createFilePath():
    wetChatDir = os.getcwd() + "\\wechat\\"
    if not os.path.exists(wetChatDir):
        os.mkdir(wetChatDir)
    return wetChatDir

# 保存好友头像
def saveWetChatDir(wetChatDir):
    bot = Bot()# 初始化机器人，扫码登陆
    friends = bot.friends(update=True)
    num = 0
    for friend in friends:
        friend.get_avatar(wetChatDir + '\\' + str(num) + ".jpg")
        print('好友昵称:%s' % friend.nick_name)
        num = num + 1
        
def jointWetChatDir(path):
    # 获取文件夹内头像个数
    length = len(os.listdir(path))
    # 设置画布大小
    image_size = 2560
    # 设置每个头像大小
    each_size = math.ceil(2560 / math.floor(math.sqrt(length)))
    # 计算所需各行列的头像数量
    x_lines = math.ceil(math.sqrt(length))
    y_lines = math.ceil(math.sqrt(length))
    image = Image.new('RGB', (each_size * x_lines, each_size * y_lines))
    x = 0
    y = 0
    for (root, dirs, files) in os.walk(path):
        for pic_name in files:
            # 增加头像读取不出来的异常处理
                try:
                    with Image.open(path + pic_name) as img:
                        img = img.resize((each_size, each_size))
                        image.paste(img, (x * each_size, y * each_size))
                        x += 1
                        if x == x_lines:
                            x = 0
                            y += 1
                except IOError:
                    print("头像读取失败")
    img = image.save(os.getcwd() + "/wetchat.png")
    print('微信好友头像拼接完成!')
    
if __name__ == '__main__':
    wetChatDir = creat_filepath()
    saveWetChatDir(wetChatDir)
    jointWetChatDir(wetChatDir)

结果如下

生成wechat文件夹，里面会下载你所有的好友图片。同时也会生成一张wetchat.png即为拼接结果。好友超过2000，可能会犯密集恐惧症。图片超过20M了，放张截图吧。

对微信好友男女性别，省份和城市分析

直接上python如下：

import itchat
import pandas as pd
from pyecharts import Pie, Map, Style, Page, Bar

# 根据key值得到对应的信息
def get_key_info(friends_info, key):
    return list(map(lambda friend_info: friend_info.get(key), friends_info))

# 获得所需的微信好友信息
def get_friends_info():
    itchat.auto_login(hotReload=True)
    friends = itchat.get_friends()
    friends_info = dict(
        # 省份
        province = get_key_info(friends, "Province"),
        # 城市
        city = get_key_info(friends, "City"),
        # 昵称
        nickname = get_key_info(friends, "Nickname"),
        # 性别
        sex = get_key_info(friends, "Sex"),
        # 签名
        signature = get_key_info(friends, "Signature"),
        # 备注
        remarkname = get_key_info(friends, "RemarkName"),
        # 用户名拼音全拼
        pyquanpin = get_key_info(friends, "PYQuanPin")
    )
    return friends_info
    
# 性别分析
def analysisSex():
    friends_info = get_friends_info()
    df = pd.DataFrame(friends_info)
    sex_count = df.groupby(['sex'], as_index=True)['sex'].count()
    temp = dict(zip(list(sex_count.index), list(sex_count)))
    data = {}
    data['保密'] = temp.pop(0)
    data['男'] = temp.pop(1)
    data['女'] = temp.pop(2)
    # 画图
    page = Page()
    attr, value = data.keys(), data.values()
    chart = Pie('微信好友性别比')
    chart.add('', attr, value, center=[50, 50],redius=[30, 70], is_label_show=True,legend_orient='horizontal',legend_pos='center',legend_top='bottom', is_area_show=True)
    page.add(chart)
    page.render('analysisSex.html')
# 省份分析
def analysisProvince():
    #得到朋友信息
    friends_info = get_friends_info()
    #存到pd元组中
    df = pd.DataFrame(friends_info)
    #省排序
    province_count = df.groupby('province', as_index=True)['province'].count().sort_values()
    temp = list(map(lambda x: x if x != '' else '未知',list(province_count.index)))
    # 画图
    page = Page()
    style = Style(width=1100, height=600)
    style_middle = Style(width=900, height=500)
    attr, value = temp, list(province_count)
    chart1 = Map('好友分布(中国地图)', **style.init_style)
    chart1.add('', attr, value, is_label_show=True, is_visualmap=True, visual_text_color='#000')
    page.add(chart1)
    chart2 = Bar('好友分布柱状图', **style_middle.init_style)
    chart2.add('', attr, value, is_stack=True, is_convert=True,
               label_pos='inside', is_legend_show=True, is_label_show=True)
    page.add(chart2)
    page.render('analysisProvince.html')
   
# 具体省份分析
def analysisCity(province):
    friends_info = get_friends_info()
    df = pd.DataFrame(friends_info)
    temp1 = df.query('province == "%s"' % province)
    city_count = temp1.groupby('city', as_index=True)['city'].count().sort_values()
    attr = list(map(lambda x: '%s市' % x if x != '' else '未知', list(city_count.index)))
    value = list(city_count)
    # 画图
    page = Page()
    style = Style(width=1100, height=600)
    style_middle = Style(width=900, height=500)
    chart1 = Map('%s好友分布' % province, **style.init_style)
    chart1.add('', attr, value, maptype='%s' % province, is_label_show=True,
               is_visualmap=True, visual_text_color='#000')
    page.add(chart1)
    chart2 = Bar('%s好友分布柱状图' % province, **style_middle.init_style)
    chart2.add('', attr, value, is_stack=True, is_convert=True, label_pos='inside', is_label_show=True)
    page.add(chart2)
    page.render('analysisCity.html')

#主函数
if __name__ == '__main__':
    analysisSex()
    analysisProvince()
    analysisCity("湖南")

###结果如下
1.性别统计分析结果
2.好友省份分布统计