hive 同一个表不同分区的数据求和

业务表如下：

hive> desc sc_kekaoxing;
OK
id                      string                  可靠小表id              
car_id                  string                  car表id              
month                   string                  月份                  
name                    string                  可靠名称父级              
count                   string                  可靠数量                
dt                      string                                      
dp                      string                                      

# Partition Information          
# col_name              data_type               comment             

dt                      string                                      
dp                      string                                      
Time taken: 0.128 seconds, Fetched: 13 row(s)

hive> select * from sc_kekaoxing limit 5;
OK
1       10      201612  车身外观        5       201612  qiche
1       11      201612  车身外观        6       201612  qiche
1       12      201612  车身外观        7       201612  qiche
1       13      201612  车身外观        2       201612  qiche
1       14      201612  车身外观        17      201612  qiche
Time taken: 0.386 seconds, Fetched: 5 row(s)

业务需求：不同dt分区下的字段count的值需要累加：

INSERT OVERWRITE local directory '/tmp/t01'
row format delimited fields terminated by '\t'
select 
if(aa.id is null,bb.id,aa.id),
if(aa.car_id is null,bb.car_id,aa.car_id),
'201702',
if(aa.name is null,bb.name,aa.name),
case 
when aa.count is not null and bb.count is not null 
  then aa.count+bb.count 
when aa.count is null and bb.count is not null 
  then bb.count 
when aa.count is not null and bb.count is null 
  then aa.count
else 0 end
from 
  (select * from sc_kekaoxing where dt='201612') aa 
full outer join 
  (select * from sc_kekaoxing where dt='201702') bb 
on aa.id=bb.id and aa.car_id=bb.car_id and aa.name=bb.name;