17年使用账，需要调整一个product_no，读取数据待优化

审计第二阶段/白条（非去哪儿）/白条在贷明细--按order_no维度.py

@@ -16,8 +16,8 @@ engine_new_transaction = create_engine(
 
 #假如导出10月份的数据，则end_point 是11月份第一天，watch_point 是10月份最后一天
 
-end_point = '2017-11-01'
-watch_point = datetime.date(2017, 10, 31)
+end_point = '2017-10-01'
+watch_point = datetime.date(2017, 9, 30)
 
 out_file_path = u'E:/审计出表/11月份开始使用表数据/11月份白条出表/在贷/'
 query_loan = """
@@ -209,6 +209,7 @@ df_ali = pd.read_sql(repay_ali % end_point, engine_new_transaction)
 df_bank = pd.read_sql(repay_bank % end_point, engine_new_transaction)
 df_cash = pd.read_sql(repay_cash % (end_point, end_point), engine_new_transaction)
 df_repay = pd.concat([df_online, df_ali, df_bank, df_cash], axis=0, ignore_index=True)
+
 df_repay.dropna(axis=0, inplace=True)
 df_loan = df_loan.loc[df_loan['principle'] > 0]
 df_info.order_no = df_info.order_no.astype(str)
@@ -220,10 +221,9 @@ df_repay = df_repay.groupby(['order_no', 'term_no'])['paid_principle'].agg({'sum
 df_out = pd.merge(df_loan, df_repay, on=['order_no', 'term_no'], how='left')
 df_out['deadline'] = pd.to_datetime(df_out.deadline)
 df_out.fillna(0, inplace=True)
-
+# watch_point = datetime.date(2017, 9, 30)
 df_out['flag'] = (watch_point-df_out['deadline']).dt.days
-df_out=df_out[df_out['flag']>0]
-df_out=df_out[df_out['zaidai']>0]
+
 df_out['zaidai'] = df_out['principle'] - df_out['paid_principle']
 df_out.loc[df_out['order_no'].isin(
     ['trainmall1483695024282', 'trainmall1484301600672', 'trainmall1486362269554', 'trainmall1487920296689',
@@ -231,7 +231,12 @@ df_out.loc[df_out['order_no'].isin(
      'trainmall1484301600672', 'trainmall1486362269554', 'grabtrainmall1484650405321', 'trainmall1495785791995',
      'trainmall1499935282194', 'trainmall16008368458050561', 'trainmall16008408858634241',
      'trainmall16112348780571649', 'trainmall16112552463312897']), 'zaidai'] = 0
+#只在17年使用的部分，调账使用
+'''
+df_out.loc[df_out['product_no']=='300156290067001502503609','flag'] = 1
+'''
 
+df_out.loc[(df_out['flag']<=0),'0'] = df_out['zaidai']
 df_out.loc[(df_out['flag']>=1) &(df_out['flag']<=30),'01-30'] = df_out['zaidai']
 df_out.loc[(df_out['flag']>=31) &(df_out['flag']<=60),'31-60'] = df_out['zaidai']
 df_out.loc[(df_out['flag']>=61) &(df_out['flag']<=90),'61-90'] = df_out['zaidai']
@@ -242,7 +247,7 @@ df_out.loc[(df_out['flag']>=181) &(df_out['flag']<=360),'181-360'] = df_out['zai
 df_out.loc[(df_out['flag']>=361),'361+'] = df_out['zaidai']
 
 # df_out['zaidai'] = df_out['zaidai'].apply(lambda x: np.round(x, 2))
-df = df_out.groupby(['order_no'])['01-30','31-60','61-90','91-120','121-150','151-180','181-360','361+'].sum().reset_index()
+df = df_out.groupby(['order_no'])['0','01-30','31-60','61-90','91-120','121-150','151-180','181-360','361+'].sum().reset_index()
 df.fillna(0,inplace=True)
 df_loan.order_no = df_loan.order_no.astype(str)
 df = pd.merge(df, df_info, on='order_no', how='left')
@@ -250,7 +255,7 @@ df = pd.merge(df, df_info, on='order_no', how='left')
 df.rename(columns={'order_no': u'订单号'
                    }, inplace=True)
 
-df = df[[u'商户', u'订单号', u'订单金额', u'合同期数', u'放款时间','01-30','31-60','61-90','91-120','121-150','151-180','181-360','361+']]
+df = df[[u'商户', u'订单号', u'订单金额', u'合同期数', u'放款时间','0','01-30','31-60','61-90','91-120','121-150','151-180','181-360','361+']]
 
 all_wb = pyexcelerate.Workbook()
 for i in xrange(0, len(df), max_limit):