oracle 按照时间分组统计求和

select t.year,
 
t.季度 时间类型,
 
sum(A)over (order by rownum rows BETWEEN unbounded preceding and current row)A,
sum(B)over (order by rownum rows BETWEEN unbounded preceding and current row)B,
sum(C)
over (order by rownum rows BETWEEN
unbounded preceding and current row)
C ,
sum(D)
over (order by rownum
rows BETWEEN unbounded preceding and current row)
D ,
sum(升级)
over (order
by row
4000
num rows BETWEEN unbounded preceding and current row)
升级 ,
sum(降级)
over (order by rownum rows BETWEEN unbounded preceding and
current row)
降级,
sum(正常受理)
over (order by rownum rows BETWEEN unbounded
preceding and current row)
正常受理 ,
sum(正常办结)
over (order by rownum rows
BETWEEN unbounded preceding and current row)
正常办结,
sum(挽回经济损失)
over (order by rownum rows BETWEEN unbounded preceding and current
row)
挽回经济损失，

   sum(工商)over (order by rownum rows BETWEEN unbounded preceding and current row)工商,

sum(质监)over (order by rownum rows BETWEEN unbounded preceding and current row)质监,

sum(食药监)over (order by rownum rows BETWEEN unbounded preceding and current row)食药监,

sum(物价)over (order by rownum rows BETWEEN unbounded preceding and current row)物价
from (
select
t.year ,
 
t.季度,
 
sum(decode(t.F_CREDIT_LEVEL, 'A', t.NUM , 0)) "A",
sum(decode(t.F_CREDIT_LEVEL, 'B', t.NUM , 0)) "B",
sum(decode(t.F_CREDIT_LEVEL, 'C', t.NUM , 0)) "C",
sum(decode(t.F_CREDIT_LEVEL, 'D', t.NUM , 0)) "D",
sum(decode(t.F_CREDIT_LEVEL, '升级', t.NUM , 0)) "升级",
sum(decode(t.F_CREDIT_LEVEL, '降级', t.NUM , 0)) "降级",
sum(decode(t.F_CREDIT_LEVEL, '已受理', t.NUM , 0)) "正常受理",
sum(decode(t.F_CREDIT_LEVEL, '已办结', t.NUM , 0))"正常办结",
sum(decode(t.F_CREDIT_LEVEL, '挽回经济损失', t.NUM , 0)) "挽回经济损失",

sum(decode(t.F_CREDIT_LEVEL, '工商', t.NUM , 0)) "工商",

sum(decode(t.F_CREDIT_LEVEL, '质监', t.NUM , 0)) "质监",

sum(decode(t.F_CREDIT_LEVEL, '食药监', t.NUM , 0)) "食药监",

sum(decode(t.F_CREDIT_LEVEL, '物价', t.NUM , 0)) "物价"
from
(
select
"TO_CHAR"(F_FOUND_DATE,'yyyy')
year,
 
 
to_char(F_FOUND_DATE,'MM') 季度,
 
F_CREDIT_LEVEL,count(*) num
FROM(SELECT F_FOUND_DATE,F_CREDIT_LEVEL
from KEY_ENTITY_INFO union all
SELECT to_date(F_CANCEL_DATE,'yyyy-mm-dd
hh24:mi:ss'),F_CREDIT_LEVEL
from KEY_ENTITY_CANCELLED where
F_CREDIT_LEVEL !='Z')
where F_FOUND_DATE>"TO_DATE"('2014-01-01',
'yyyy-MM-dd')
group by
"TO_CHAR"(F_FOUND_DATE,'yyyy')
 
 
,to_char(F_FOUND_DATE,'MM')
 
,F_CREDIT_LEVEL
union all
select "TO_CHAR"(F_UPDATE_DATE,'yyyy')
year,
 
 
to_char(F_UPDATE_DATE,'MM') 季度,
 
'升级',count(*) num
FROM KEY_CREDIT_HISTORY
where
F_UPDATE_DATE>"TO_DATE"('2014-01-01', 'yyyy-MM-dd')
and
F_LEVEL_BEFORE>F_LEVEL_AFTER
group by "TO_CHAR"(F_UPDATE_DATE,'yyyy')
 
 
,to_char(F_UPDATE_DATE,'MM')
 
union all
select "TO_CHAR"(F_UPDATE_DATE,'yyyy')
year,
 
 
to_char(F_UPDATE_DATE,'MM') 季度,
 
'降级',count(*) num
FROM KEY_CREDIT_HISTORY
where
F_UPDATE_DATE>"TO_DATE"('2014-01-01', 'yyyy-MM-dd')
and
F_LEVEL_BEFORE<F_LEVEL_AFTER
group by
"TO_CHAR"(F_UPDATE_DATE,'yyyy')
 
 
,to_char(F_UPDATE_DATE,'MM')
 
union all
SELECT
'2013',
 
 
to_char(F_FOUND_DATE,'MM') 季度,
 
F_CREDIT_LEVEL,COUNT (*) num
FROM
(SELECT
F_FOUND_DATE,
F_CREDIT_LEVEL
FROM
KEY_ENTITY_INFO
UNION ALL
SELECT
TO_DATE (
F_CANCEL_DATE,
'yyyy-mm-dd
hh24:mi:ss'
),

F_CREDIT_LEVEL
FROM
KEY_ENTITY_CANCELLED
WHERE
F_CREDIT_LEVEL != 'Z'
)
WHERE
F_FOUND_DATE
< "TO_DATE" ('2014-01-01',
'yyyy-MM-dd')
GROUP BY
F_CREDIT_LEVEL
 
 
,to_char(F_FOUND_DATE,'MM')
 
UNION ALL
SELECT
'2013',
 
 
to_char(F_UPDATE_DATE,'MM') 季度,
 
'升级',
COUNT (*) num
FROM
KEY_CREDIT_HISTORY
WHERE
F_UPDATE_DATE <
"TO_DATE" ('2014-01-01', 'yyyy-MM-dd')
AND F_LEVEL_BEFORE >
F_LEVEL_AFTER
 
 
group by to_char(F_UPDATE_DATE,'MM')
 
UNION ALL
SELECT
'2013',
 
 
to_char(F_UPDATE_DATE,'MM') 季度,
 
'降级',
COUNT (*) num
FROM
KEY_CREDIT_HISTORY
WHERE
F_UPDATE_DATE
<"TO_DATE" ('2014-01-01', 'yyyy-MM-dd')
AND F_LEVEL_BEFORE <
F_LEVEL_AFTER
 
 
group by to_char(F_UPDATE_DATE,'MM')
 

union all
select
"TO_CHAR"(F_REG_TIME,'yyyy')year,
 
 
to_char(F_REG_TIME,'MM') 季度,
 
'已受理' type,
count(*) num
FROM
BIZ_COMPLAINTS_HISTORY
group by
"TO_CHAR"(F_REG_TIME,'yyyy')
 
 
,to_char(F_REG_TIME,'MM')
 
union all
select
"TO_CHAR"(F_CLOSE_TIME,'yyyy')year,
 
 
to_char(F_CLOSE_TIME,'MM') 季度,
 
'已办结' type,
count(*) num
FROM
BIZ_COMPLAINTS_HISTORY
group by
"TO_CHAR"(F_CLOSE_TIME,'yyyy')
 
 
,to_char(F_CLOSE_TIME,'MM')
 
union all
select
"TO_CHAR"(F_REG_TIME,'yyyy')year,
 
 
to_char(F_REG_TIME,'MM'),
 
'挽回经济损失' type,
sum(F_RECOUPEDECONOMY_LOST)
FROM
BIZ_COMPLAINTS_HISTORY
group by
"TO_CHAR"(F_REG_TIME,'yyyy')
 
 
,to_char(F_REG_TIME,'MM')
 

union all 

    select
"TO_CHAR"(F_SJ_CJSJ,'yyyy')year,
 
 
to_char(F_SJ_CJSJ,'MM'),
 
 F_AY_AJLY,
count(*)
FROM
(

SELECT
CASE
WHEN T .F_AY_AJLY = 1 THEN
'工商'
WHEN T .F_AY_AJLY = 2 THEN
'质监'
WHEN T .F_AY_AJLY = 3 THEN
'食药监'
WHEN T .F_AY_AJLY = 4 THEN
'物价'
ELSE
''
END F_AY_AJLY,
t2.F_SJ_CJSJ

FROM
(
SELECT
F_Id,
SUBSTR (
T .ca,
INSTR (T .ca, ',', 1, c.lv) + 1,
INSTR (T .ca, ',', 1, c.lv + 1) - (INSTR(T .ca, ',', 1, c.lv) + 1)
) AS F_AY_AJLY
FROM
(
SELECT
F_Id,
',' || F_AY_AJLY || ',' AS ca,
LENGTH (F_AY_AJLY || ',') - NVL (
LENGTH (REPLACE(F_AY_AJLY, ',')),
0
) AS cnt,
F_SJ_CJSJ
FROM
act_aj_ayxx
WHERE
1 = 1
) T,
(
SELECT
LEVEL lv
FROM
dual CONNECT BY LEVEL <= 5
) c
WHERE
c.lv <= T .cnt
) T

LEFT JOIN ACT_AJ_AYXX t2 ON T .f_id = t2.f_id

   )
group by
"TO_CHAR"(F_SJ_CJSJ,'yyyy'),

to_char(F_SJ_CJSJ,'MM'),

    F_AY_AJLY

)t
group by t.year
 
,t.季度
 
ORDER BY year
 
,t.季度
 
)
t
group by rownum,t.year,
 
t.季度,
 
t.A,t.B,t.C,T.D,t.升级,t.降级,T.正常受理,t.正常办结,t.挽回经济损失,t.工商,t.质监,t.食药监,t.物价
order by rownum