data-verification/Shihui/customer_count.py

# encoding=utf8

import pymysql
import pandas as pd
import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt
from io import BytesIO
from datetime import datetime, date
from datetime import timedelta
import  xlsxwriter  

DATE = None
STEP = None

today = datetime.today()
year = today.year
month = today.month 
if month == 1:
    month = 12
    year -= 1
else:
    month -= 1 
date1 = datetime(year, month, 1)
date2 = datetime(today.year, today.month, 1)
date1_week = date1.isocalendar()[1]
date2_week = date2.isocalendar()[1]
if DATE is None:
    DATE = date1
if STEP is None:
    STEP = (date2 - date1).days



def get_data():
  mysql_host = '47.95.217.180'
  mysql_port = 3306
  mysql_user = 'rtech'
  mysql_password = 'S5RE4dQ65MphYk7F'
  mysql_db = 'fmp'

  db = pymysql.connect(host=mysql_host,port=mysql_port,user=mysql_user, password=mysql_password,db=mysql_db,charset = 'utf8')
  print("数据库连接成功")

  df = pd.read_sql(sql,con=db)
  db.close
  
  mpl.rcParams["font.sans-serif"] = ["SimHei"]
  mpl.rcParams["axes.unicode_minus"] = False
  

  df['enterprise_id'] = 'shihuibeijing'
  temp_actula_price = df['flow_amount']
  total_flow_amount = df['flow_amount'].sum(axis = 0)
  print(total_flow_amount)
 
  df['status'] = df['status'].str.replace('1','等待充值')
  df['status'] = df['status'].str.replace('2','充值中')
  df['status'] = df['status'].str.replace('4','充值失败')
  df['status'] = df['status'].str.replace('6','充值成功')
 
  temp_callback_status = df['status']
 
  '''
  df['channel_id'] = df['channel_id'].str.replace('104','亚杉移动网厅')
  df['channel_id'] = df['channel_id'].str.replace('117','亚杉电信网厅')
  df['channel_id'] = df['channel_id'].str.replace('92','人工充值移动')
  df['channel_id'] = df['channel_id'].str.replace('93','人工充值联通')
  df['channel_id'] = df['channel_id'].str.replace('94','人工充值电信')
  df['channel_id'] = df['channel_id'].str.replace('89','兆蓉移动')
  df['channel_id'] = df['channel_id'].str.replace('90','兆蓉联通')
  df['channel_id'] = df['channel_id'].str.replace('91','兆蓉电信')
  df['channel_id'] = df['channel_id'].str.replace('95','兆蓉移动网厅')
  df['channel_id'] = df['channel_id'].str.replace('116','兆蓉联通网厅')
  df['channel_id'] = df['channel_id'].str.replace('105','兆蓉电信网厅')
  df['channel_id'] = df['channel_id'].str.replace('83','满帆移动网厅')
  df['channel_id'] = df['channel_id'].str.replace('84','满帆联通网厅')
  df['channel_id'] = df['channel_id'].str.replace('85','满帆电信网厅')
  df['channel_id'] = df['channel_id'].str.replace('96','自动补单移动通道')
  df['channel_id'] = df['channel_id'].str.replace('97','自动补单联通通道')
  df['channel_id'] = df['channel_id'].str.replace('98','自动补单电信通道')
  df['channel_id'] = df['channel_id'].str.replace('113','自动补单移动通道2')
  df['channel_id'] = df['channel_id'].str.replace('114','自动补单联通通道2')  
  df['channel_id'] = df['channel_id'].str.replace('115','自动补单电信通道2')
  df['channel_id'] = df['channel_id'].str.replace('118','智信全国移动')
  df['channel_id'] = df['channel_id'].str.replace('119','智信全国联通')
  df['channel_id'] = df['channel_id'].str.replace('120','智信全国电信')
  df['channel_id'] = df['channel_id'].str.replace('121','易迅捷移动网厅')
  df['channel_id'] = df['channel_id'].str.replace('122','易迅捷电信网厅')
  df['channel_id'] = df['channel_id'].str.replace('123','易迅捷联通网厅')
  df['channel_id'] = df['channel_id'].str.replace('124','亚杉联通网厅')
  '''
  
  
  province = [] #省份
  provider = [] #运营商
  for area in df['mobile_home']:
      province.append(area.split('-',2)[0])
      provider.append(area.split('-',2)[2])
      #print(area)
      
  '''
  col_name = df.columns.tolist()
  col_name.insert(col_name.index('used_mobile'),'actual_price')
  col_name.insert(col_name.index('mobile_home')+1,'area')
  col_name.insert(col_name.index('status')+1,'callback_status')
  df.reindex(columns=col_name)
  '''
  
  df = df.reindex(columns=['order_id','enterprise_id','flow_amount','actual_price','used_mobile','mobile_home','area',
                      'apply_date','check_time','status','use_time','callback_status','channel_name'])

  df['order_id'] = df['order_id'].apply(str)
  df['actual_price'] = temp_actula_price
  df['callback_status'] = temp_callback_status
  df['area'] = pd.Series(province)
  df['mobile_home'] = pd.Series(provider)
  #print(df['mobile_home'])
  
  provider_order =  df['mobile_home'].value_counts() #运营商订单数
  provider_money = df.groupby('mobile_home').agg({'flow_amount':'sum'})#运行商面额
  
  periods = pd.PeriodIndex(df['apply_date'],freq='D')
  df['datetime'] = periods
  every_money = df.groupby('datetime').agg({'flow_amount':'sum'}) #每日充值面额
  #print(provider_money)
  #print(every_money['flow_amount'].values) 
 
  pay_status = df['status'].value_counts() #充值状态占比
  
  pay_price = df.groupby('flow_amount').agg({'flow_amount':'sum'})#充值面额占比
  
  pay_area = df['area'].value_counts().sort_values(ascending=False)#充值地域占比  
  #print(pay_price)
 
  
  
  pay_time = []
  
  '''
  for i in df['use_time'].values:
    if i/60 < 10:
        pay_time.append('0-10分钟')
    elif 10 <= i/60 < 20:
        pay_time.append('10-20分钟')
    elif 20 <= i/60 < 30:
        pay_time.append('20-30分钟')
    elif 30 <= i/60 < 40:
        pay_time.append('30-40分钟')
    elif 40 <= i/60 < 50:
        pay_time.append('40-50分钟')
    elif 50 <= i/60 <= 60:
        pay_time.append('50-60分钟')
    else :
        pay_time.append('60分钟以上')
        #pass
  '''  
  
  for i in df['use_time'].values:
    if i/60 < 20:
        pay_time.append('0-10分钟')
    elif 20 <= i/60 < 30:
        pay_time.append('10-20分钟')
    elif 30 <= i/60 < 40:
        pay_time.append('20-30分钟')
    elif 40 <= i/60 < 50:
        pay_time.append('30-40分钟')
    elif 50 <= i/60 < 60:
        pay_time.append('40分钟以上')
    else :
        #pay_time.append('60分钟以上')
        pass      
  
  use_time_count = pd.value_counts(pay_time).sort_index()#充值时长占比
  #use_time_count = pd.value_counts(pay_time)#充值时长占比
  use_df = pd.DataFrame(use_time_count)
  
  #print(use_time_count)
  #print(type(use_time_count))
  #print(use_df)
  
  use_df = use_df.div(use_df.sum(axis=0), axis=1)
  
  #print(a[0].values[0])
  
  
  df.drop(['datetime'],axis=1,inplace=True)

  
   
  
  df.columns = ['订单号','应用id','充值面额','实付金额','手机号','运营商','归属地','发送日期','回调日期',
                '订单状态','用时（秒）','回调状态','通道名称']
  #UPLOAD_FOLDER = './count_data/北京十荟每月数据统计.xlsx'
  #print(222)
  #UPLOAD_FOLDER = './北京十荟每月数据统计.csv'
  UPLOAD_FOLDER = './北京十荟{}月数据统计.xlsx'.format(pre_month.month)
  print(111)
 
  
  
  writer = pd.ExcelWriter(UPLOAD_FOLDER, engine='xlsxwriter')
  #df.to_excel(writer,sheet_name=u'数据统计',index = False,encoding='utf_8_sig')
  workbook = writer.book
  sheet2 = workbook.add_worksheet(u'图表统计')
  
  #设置格式
  style1 =  workbook.add_format({'fg_color':'#FFFFFF'}) #背景色
  style2 =  workbook.add_format({'top':2,'fg_color':'#FFFFFF'}) #上边框
  style3 =  workbook.add_format({'left':2,'fg_color':'#FFFFFF'}) #左边框
  style4 =  workbook.add_format({'top':2,'left':2,'fg_color':'#FFFFFF'}) 
  style5 =  workbook.add_format({'font_size':25,'bold':True,'align':'center','valign':'vcenter','fg_color':'#20B2AA'}) #

  for row in range(150):
      if(row<3):
        sheet2.set_row(row,25,style1)
      else:
        sheet2.set_row(row,None,style1)    
  
  
  for col in range(1,22):
      sheet2.write(1,col,'',style2) 
      sheet2.write(128,col,'',style2) 
      
  for row in range(1,128):
      sheet2.write(row,1,'',style3) 
      sheet2.write(row,22,'',style3) 
  sheet2.write(1,1,'',style4)    
  
  sheet2.merge_range('B2:V3', '数据月报--十荟团', style5)

  
 
  
  #绘制本月GMV折线图
  fig = plt.figure(figsize=(10,5),dpi=200)
  print(pre_month.day)
  #sleep(5)
  x = range(1,pre_month.day+1)
  y = every_money['flow_amount'].values
  plt.plot(x,y)
  
  _xtick_labels = ['{}月{}日'.format(pre_month.month,i) for i in range(1,pre_month.day+1)]
  plt.xticks(x,_xtick_labels,rotation=45)
  plt.yticks(list(y)[::6],list(y)[::6])
  plt.xlabel('时间')
  plt.ylabel('金额 单位(元)') 
  plt.grid(alpha=0.1)#绘制网格 
  #plt.legend(loc='upper left') #图例
  plt.title('本月GMV总数为 {} 元'.format(total_flow_amount))
  plt.savefig('every_money.png')
  
  
  #绘制运营商订单占比饼图
  fig = plt.figure(figsize=(2.5,2.5),dpi=180)
  labels = list(provider_order.index)
  #print(labels)
  #imagedata=BytesIO()
  colors = ['#1E90FF','#008000','#FFA500']
  plt.pie(provider_order,labels=labels,colors=colors,autopct='%1.0f%%', shadow=False, startangle=90)
  
  plt.xticks(provider_order)
  plt.axis('equal')
  #plt.legend(loc=2)
  plt.title('运营商订单比例')
  plt.savefig('provider_order.png')
  
  
  #绘制充值时长柱状图
  fig = plt.figure(figsize=(6,6),dpi=180)
  a = list(use_time_count.index)
  _b = list(use_df[0].values)
  #print(_b)
  b = list(use_time_count.values)
  plt.bar(range(len(a)),b,width=0.6)
  plt.xticks(range(len(a)),a,rotation=45)
  #count=0
  for x,y in enumerate(_b):
    #print(x)
    #print(y)
    plt.text(x, b[x]+200, '{:.2f}'.format(y * 100) + '%', ha='center', va= 'bottom')
    #count += 1
  plt.xlabel('充值时长')
  #plt.ylabel('订单数量')
  plt.grid(alpha=0.1)
  plt.title('充值时长占比')
  plt.savefig('use_time.png')
  
  
  #绘制运营商成交面额占比饼图
  fig = plt.figure(figsize=(2.5,2.5),dpi=180)
  labels_money = list(provider_money.index)
  #print(labels_money)
  colors = ['#1E90FF','#008000','#FFA500']
  plt.pie(provider_money['flow_amount'].values,labels=labels_money,colors=colors,autopct='%1.2f%%', shadow=False,startangle=90) 
  plt.xticks(provider_money['flow_amount'].values)
  plt.axis('equal')
  #plt.legend()
  plt.title('运营商成交面额比例')
  plt.savefig('provider_money.png')
  
  #绘制充值状态占比饼图
  fig = plt.figure(figsize=(5,3.2),dpi=180)
  labels = list(pay_status.index)
  #explode = (0.3,0,0)
  #plt.pie(pay_status.values,explode =explode,labels=labels,autopct='%1.2f%%', shadow=False,startangle=0,pctdistance = 0.6,labeldistance=1.2) 
  plt.pie(pay_status.values,labels=labels,autopct='%1.2f%%', shadow=False,startangle=0,pctdistance = 0.6,labeldistance=1.2) 
  plt.xticks(pay_status.values)
  plt.axis('equal')
  plt.title('充值状态占比\n')
  plt.savefig('pay_status.png') 
  
  #绘制充值面额占比饼图
  fig = plt.figure(figsize=(5,3.2),dpi=180)
  _labels = list(pay_price.index)
  labels = [int(_labels[i])  for i in range(len(_labels))]
  explode = (0.2,0,0,0,0)
  #explode = (0.1,0,0,0,0.2,0.3)
  plt.pie(pay_price['flow_amount'].values,explode =explode,labels=labels,autopct='%1.2f%%', shadow=False,startangle=90,pctdistance = 0.6,labeldistance=1.1) 
  #plt.pie(pay_price['flow_amount'].values,labels=labels,autopct='%1.2f%%', shadow=False,startangle=90,pctdistance = 0.6,labeldistance=1.1) 
  plt.xticks(pay_price.values)
  plt.axis('equal')
  plt.title('充值面额占比\n')
  plt.savefig('pay_price.png') 
  
  
  #绘制充值地域柱状图
  fig = plt.figure(figsize=(12,8),dpi=180)
  a = list(pay_area.index)
  b = list(pay_area.values)
  plt.barh(range(len(a)),b,height=0.6,color='orange')  
  #plt.xticks(b[::2],b[::2])
  plt.yticks(range(len(a)),a)
  for y, x in enumerate(b):
    plt.text(x+1500, y-0.3, "%.0f" %x,ha='center', va= 'bottom')
  plt.xlabel('订单数量')
  plt.ylabel('省份')
  plt.grid(alpha=0.1)
  plt.title('地域分布')
  plt.savefig('area.png')

  
  sheet2.insert_image(5,3,'every_money.png',{'x_offset': 5, 'y_offset': 5}) 
  sheet2.insert_image(45,3,'provider_order.png',{'x_offset': 5, 'y_offset': 5}) 
  sheet2.insert_image(35,8,'use_time.png',{'x_offset': 5, 'y_offset': 5}) 
  sheet2.insert_image(45,17,'provider_money.png',{'x_offset': 5, 'y_offset': 5}) 
  sheet2.insert_image(69,3,'pay_status.png',{'x_offset': 10, 'y_offset': 5}) 
  sheet2.insert_image(69,13,'pay_price.png',{'x_offset': 5, 'y_offset': 5}) 
  sheet2.insert_image(88,3,'area.png',{'x_offset': 0, 'y_offset': 3}) 
   
  
  print(33) 
 
  writer.save()
  writer.close()
  
  #df.to_csv(UPLOAD_FOLDER,index=False,encoding='utf-8-sig')
   
  

if __name__ == '__main__':
    #today = datetime.today().replace(hour=0, minute=0, second=0, microsecond=0)
    #delta = timedelta(days=1)
    #求该月第一天
    #first_day = date(today.year, today.month, 1)
    #print(u'该月第一天:' + str(first_day))
    #前一个月最后一天
    pre_month = date2 - timedelta(days=1) 
    #print(u'前一个月最后一天:' + str(pre_month))
    #前一个月的第一天
    #first_day_of_pre_month = date(pre_month.year, pre_month.month, 1)
    #print(first_day_of_pre_month)

    time_start = str(date1)
    time_end = str(date2)
    #print(time_start)
    #print(time_end)

    start_year = today.year
    end_week = int(today.strftime("%U")) + 1
    
    #print('end_week is {}'.format(end_week))
    #print('date2_week is: {}'.format(date2_week))
    end_sql = '''SELECT order_id,enterprise_id,flow_amount,used_mobile,mobile_home,apply_date,
                          check_time,status,use_time,channel_name
                   FROM     
                    (SELECT order_id,enterprise_id,flow_amount,used_mobile,mobile_home,apply_date,check_time,status,
                       (UNIX_TIMESTAMP(check_time)- UNIX_TIMESTAMP(apply_date)) AS 'use_time',channel_id
                    FROM 
                       flow_order_info
                    WHERE     
                       apply_date BETWEEN '{}'  AND '{}' AND enterprise_id = 54 ) t1 LEFT JOIN access_channel_info
                    ON t1.channel_id = access_channel_info.channel_seq_id
                    ORDER BY apply_date '''

    temp_sql = ''' SELECT order_id,enterprise_id,flow_amount,used_mobile,mobile_home,apply_date,
                          check_time,status,use_time,channel_name
                   FROM       
                   (SELECT  order_id,enterprise_id,flow_amount,used_mobile,mobile_home,apply_date,check_time,status,
                    (UNIX_TIMESTAMP(check_time)- UNIX_TIMESTAMP(apply_date)) AS 'use_time',channel_id
                    FROM 
                       flow_order_info_{}
                    WHERE     
                       apply_date BETWEEN '{}'  AND '{}'  AND enterprise_id = 54) t1 LEFT JOIN access_channel_info
                    ON t1.channel_id = access_channel_info.channel_seq_id  '''
    sql = ''
    for i in range(date1_week,date2_week+1):
        str_table_suffix = "{0}{1:02d}".format(year,i)
        if i == date2_week:
            if (today - date2) < timedelta(days=7):
                sql += end_sql.format(time_start,time_end)
            else:    
                sql += temp_sql.format(str_table_suffix,time_start,time_end)
        else :
            sql += temp_sql.format(str_table_suffix,time_start,time_end) + 'UNION'

    #print(sql)
    
    get_data()