import itchat
import pandas
import seaborn as sns
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
import os
import TencentYoutuyun
import numpy as np
%matplotlib inline登陆微信网页版，需扫描二维码;获取friends数据，转成pandas.DataFrame格式，先导出，再导入，防止sns作图时出错
itchat.auto_login()friends_wechat = itchat.get_friends(update = True)
friends_df = pandas.DataFrame(data = friends_wechat[1:])
# 因为friends[0]得出的数据是自己，所以要[1:]切片
friends_df.to_csv('wechat_friends_info.csv',index_label='id')
friends_row = pandas.read_csv('wechat_friends_info.csv',encoding = 'gbk')字段太多，去掉不认识的、重复的
每个字段是什么意思？
friends_row.columnsIndex(['id', 'Alias', 'AppAccountFlag', 'AttrStatus', 'ChatRoomId', 'City',
'ContactFlag', 'DisplayName', 'EncryChatRoomId', 'HeadImgUrl',
'HideInputBarFlag', 'IsOwner', 'KeyWord', 'MemberCount', 'MemberList',
'NickName', 'OwnerUin', 'PYInitial', 'PYQuanPin', 'Province',
'RemarkName', 'RemarkPYInitial', 'RemarkPYQuanPin', 'Sex', 'Signature',
'SnsFlag', 'StarFriend', 'Statues', 'Uin', 'UniFriend', 'UserName',
'VerifyFlag'],
dtype='object')
friends = friends_row[['id','NickName','RemarkName','Province','City','Sex',\
'Signature','HeadImgUrl']]
friends.info()<class 'pandas.core.frame.DataFrame'>
RangeIndex: 199 entries, 0 to 198
Data columns (total 8 columns):
id 199 non-null int64
NickName 199 non-null object
RemarkName 65 non-null object
Province 164 non-null object
City 158 non-null object
Sex 199 non-null int64
Signature 145 non-null object
HeadImgUrl 199 non-null object
dtypes: int64(2), object(6)
memory usage: 12.5+ KB
# 设置饼图的倾斜角度，1：不倾斜
plt.axes(aspect=1)
plt.rcParams['font.family'] = ['SimHei']
plt.pie(friends.groupby('Sex',axis=0).count()['NickName'],
labels = ['unknown','male','female'],
autopct='%.1f%%' # 设置饼图中显示百分比
)
plt.title('微信好友性别概况');Text(0.5,1,'微信好友性别概况')
ncount = len(friends)
ax = sns.countplot(x='Sex',data=friends,order=[0,2,1])
plt.xlabel('Sex',fontsize=12)
plt.title('微信好友性别概况')
plt.rcParams['font.family'] = ['SimHei']
ax2 = ax.twinx()
ax2.yaxis.tick_left()
ax.yaxis.tick_right()
ax.yaxis.set_label_position('right')
ax2.yaxis.set_label_position('left')
ax.set_ylabel('人数')
ax2.set_ylabel('占比 %')
# 让‘人数’纵坐标的值为整数
fmt = '%.0f'
yticks = matplotlib.ticker.FormatStrFormatter( fmt )
ax.yaxis.set_major_formatter( yticks )
for p in ax.patches:
x=p.get_bbox().get_points()[:,0]
y=p.get_bbox().get_points()[1,1]
ax.annotate('{:.1f}%, {:d}人'.format(100.*y/ncount,int(y)), (x.mean(), y),
ha='center', va='bottom')
ax.yaxis.set_major_locator(matplotlib.ticker.LinearLocator(11))
# 设置纵轴数据范围
ax2.set_ylim(0,100)
ax.set_ylim(0,ncount)
ax2.yaxis.set_major_locator(matplotlib.ticker.MultipleLocator(10))
# 关闭坐标轴ax2 的横线
ax2.grid(None)
# 修改x轴上标签值 如此就不需要图例
a = ['unknown','female','male']
ax.set_xticklabels(a,fontsize=12);
上下两张图一对比，就能看出来为什么数据分析时都不推荐饼图了。
friends.groupby('Province',axis=0).count().sort_values('NickName',ascending=False)['NickName']Province
广东 86
陕西 16
四川 6
北京 6
江苏 5
重庆 5
湖北 4
上海 4
浙江 3
香港 3
山东 3
广西 2
Paris 2
辽宁 2
山西 2
湖南 1
福建 1
河南 1
Barcelona 1
California 1
安徽 1
天津 1
云南 1
Washington 1
Pennsylvania 1
Others 1
Karlsruhe 1
Idaho 1
Canberra 1
黑龙江 1
Name: NickName, dtype: int64
在BDP中使用“本周新增订单分布”作为模板，得到下图。
下载头像，保存到本地，判断头像是否为人像,
接入腾讯优图即可判断
folder = 'E:\\MySQL_data\\wechat_friends_img\\'
os.path.exists(folder)
j = -1 # i为1时，j为0，取出的是id=0的好友头像，保证图片名和id一致
for i in friends_wechat:
img = folder + '\\image{:d}.jpg'.format(j)
img_data = itchat.get_head_img(userName = i['UserName'])
if os.path.exists(img) == False :
with open(img,'wb') as file:
file.write(img_data)
j += 1appid = '*****'
secret_id = '******'
secret_key = '******'
userid = '******'
end_point = TencentYoutuyun.conf.API_YOUTU_END_POINT
youtu = TencentYoutuyun.YouTu(appid, secret_id, secret_key, userid, end_point)friends['ishuman'] = 0
# 每次用这个方法得到新列时都会出现下面这一堆说明，但是也都能成功E:\miniconda\envs\course_py35\lib\site-packages\ipykernel_launcher.py:1: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead
See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy
"""Entry point for launching an IPython kernel.
for i in friends['id']:
result = youtu.DetectFace('E:\\MySQL_data\\wechat_friends_img\\image{}.jpg'.format(i),
mode = 0, data_type = 0)
if result['face']:
friends.set_value(i,'ishuman',1)
else:
friends.set_value(i,'ishuman',0)friends['ishuman'].mean()0.38693467336683418
plt.rcParams['font.family'] = ['SimHei']
plt.title('性别与使用人像')
ax = sns.countplot(x='ishuman',hue='Sex',data=friends,order = [1,0])
ax.yaxis.tick_right()
ax.yaxis.set_label_position('right')
ax.set_ylabel('人数')
for p in ax.patches:
x=p.get_bbox().get_points()[:,0]
y=p.get_bbox().get_points()[1,1]
ax.annotate('{:d}人'.format(int(y)), (x.mean(), y),
ha='center', va='bottom')
ax.yaxis.set_major_locator(matplotlib.ticker.LinearLocator(11))
ax.set_ylim(0,80)
xtick_labels = ['使用人像','不使用人像']
ax.set_xticklabels(xtick_labels,fontsize=12);
结论：男、女在是否使用人脸作为头像上无明显差异
未注明性别的，更不愿意使用人脸作为头像
friends['issign'] = friends['Signature'].notnull()ncount = len(friends)
ax = sns.countplot(x='issign',hue = 'Sex',data=friends)
plt.rcParams['font.family'] = ['SimHei']
plt.xlabel('是否签名',fontsize=12)
plt.title('微信好友签名概况')
ax2 = ax.twinx()
ax2.yaxis.tick_left()
ax.yaxis.tick_right()
ax.yaxis.set_label_position('right')
ax2.yaxis.set_label_position('left')
ax.set_ylabel('人数')
ax2.set_ylabel('占比 %')
# 让‘人数’纵坐标的值为整数
fmt = '%.0f'
yticks = matplotlib.ticker.FormatStrFormatter( fmt )
ax.yaxis.set_major_formatter( yticks )
for p in ax.patches:
x=p.get_bbox().get_points()[:,0]
y=p.get_bbox().get_points()[1,1]
ax.annotate('{:.1f}%,{:d}人'.format(100.*y/ncount,int(y)), (x.mean(), y),
ha='center', va='bottom')
ax.yaxis.set_major_locator(matplotlib.ticker.LinearLocator(11))
# 设置纵轴数据范围
ax2.set_ylim(0,50)
ax.set_ylim(0,100)
ax2.yaxis.set_major_locator(matplotlib.ticker.MultipleLocator(10))
# 关闭坐标轴ax2 的横线
# ax2.grid(None)
# 修改x轴上标签值 如此就不需要图例
a = ['未签名','有签名']
ax.set_xticklabels(a,fontsize=12);
和男性相比，女性使用签名的比例更大
friends.to_csv('wechat_friends_data.csv',
index = False, # 因为已经有了id这列
columns = ['id','Sex','issign','ishuman','Province','City']
)