【Python数据分析】用户通话行为分析

  1 #coding=utf-8
  2 import xlrd
  3 import matplotlib.pyplot as plt
  4 import UserDataInfo
  5 
  6 def readData(url):
  7     data=xlrd.open_workbook(url)
  8     table=data.sheets()[0]  #多张sheet的情况读取第一张
  9     nrows=table.nrows
 10     ncols=table.ncols
 11     list=[]
 12     for rownum in range(0,nrows):
 13         row=table.row_values(rownum)
 14         for i in range(0,ncols):        #转码unicode转utf-8
 15             row[i]=row[i].encode(‘utf-8‘)
 16         if row:
 17             list.append(row)
 18     return list
 19 
 20 #行为分析
 21 def behavior_analysis(datalist):
 22     t=1
 23     for line in datalist:
 24         if(t==1):
 25             t=3
 26             continue
 27         dh=dateDecode(line[2])
 28         day=int(dh[0])
 29         hour=int(dh[-1])
 30 
 31         user.day_intervel[day]+=1
 32         user.time_intervel[hour/2]+=1
 33 
 34 
 35         timeStr=line[3]
 36         timelong = timeDecode(timeStr)
 37 
 38         if line[4]==‘主叫‘:
 39             user.calling_times+=1
 40             user.calling_long+=timelong
 41         if line[4]==‘被叫‘:
 42             user.called_times+=1
 43             user.called_long+=timelong
 44 
 45 
 46     user.call_times=user.calling_times+user.called_times    #总次数
 47     user.call_long=user.calling_long+user.called_long       #总时长
 48 
 49 #解码时间
 50 def timeDecode(timeStr):
 51     hour = min = sec = 0
 52     if timeStr.find(‘小时‘) != -1:
 53         hour = timeStr.split(‘小时‘)[0]
 54         timeStr = timeStr.split(‘小时‘)[1]
 55     if timeStr.find(‘分‘) != -1:
 56         min = timeStr.split(‘分‘)[0]
 57         timeStr = timeStr.split(‘分‘)[1]
 58     if timeStr.find(‘秒‘) != -1:
 59         sec = timeStr.split(‘秒‘)[0]
 60 
 61     timelong = int(sec) + int(min) * 60 + int(hour) * 60 * 60
 62     return timelong
 63 
 64 #编码时间
 65 def timeEncode(time):
 66     strtime=str(time%60)+"秒"
 67     time/=60
 68     if(time!=0):
 69         strtime=str(time%60)+"分"+strtime
 70         time/=60
 71         if (time != 0):
 72             strtime = str(time % 60) + "小时"+strtime
 73     return strtime
 74 
 75 #解码日期
 76 def dateDecode(dateStr):
 77     dateStrlist=dateStr.split(‘ ‘)
 78     dayStr=dateStrlist[0]
 79     dayStrlist=dayStr.split(‘-‘)
 80     day=dayStrlist[-1]
 81 
 82     timeStr=dateStrlist[-1]
 83     timeStrlist=timeStr.split(‘:‘)
 84     timeHour=timeStrlist[0]
 85 
 86     dayAndHour=[]
 87     dayAndHour.append(day)
 88     dayAndHour.append(timeHour)
 89     return dayAndHour
 90 
 91 
 92 def printout():
 93     print "被叫次数：", user.called_times
 94     print "被叫时长：", timeEncode(user.called_long)
 95 
 96     print "主叫次数：", user.calling_times
 97     print "主叫时长：", timeEncode(user.calling_long)
 98 
 99     print "总次数：",user.call_times
100     print "总时长：",timeEncode(user.call_long)
101 
102     print "日期",user.day_intervel
103     print "时段",user.time_intervel
104 
105 #数据可视化
106 def dataVisualization(userinfo):
107     plt.plot(userinfo.day_intervel, ‘k‘)
108     plt.plot(userinfo.day_intervel, ‘bo‘)
109     plt.xlabel(u‘日          期‘)
110     plt.ylabel(u‘通话次数‘)
111     plt.title(u‘每日通话分析‘)
112     plt.grid(color=‘#95a5a6‘, linestyle=‘--‘, linewidth=1, axis=‘y‘, alpha=0.4)
113     plt.show()
114 
115 url="/Users/SeeKHit/Downloads/2017年01月语音通信.xls"
116 datalist=readData(url)
117 
118 user=UserDataInfo.UserDataInfo()
119 behavior_analysis(datalist)
120 
121 printout()
122 dataVisualization(user)

 1 #coding=utf-8
 2 # 用户通话信息类记录
 3 class UserDataInfo(object):
 4     def __init__(self):
 5         self.calling_long = 0   #主叫时长
 6         self.called_long = 0    #被叫时长
 7         self.call_long = 0      #总时长
 8 
 9         self.calling_times = 0  #主叫次数
10         self.called_times = 0   #被叫次数
11         self.call_times = 0     #总次数
12 
13         self.time_intervel=[]   #通话时段
14         for i in range(0,12):
15             self.time_intervel.append(0)
16 
17         self.day_intervel=[]    #每日通话次数
18         for i in range(0,31):
19             self.day_intervel.append(0)
20 
21         self.call_freq= None       #通话最频繁