如何UNPIVOT到多个列？
我有一张如下图所示的图片
我需要输出如下图像
我有一个使用union查询所需输出的查询
select distinct cd_5 as ORG_UNIT_CD, upper('"'||nm_5||'"') as ORG_UNIT_NM, cd4 as PARENT_ORG_UNIT_CD, 5 as ORG_UNIT_LEVEL
from TBL
where cd_5 <> 0
union
select distinct cd_4 as ORG_UNIT_CD, upper('"'||nm_4||'"') as ORG_UNIT_NM, cd_3 as PARENT_ORG_UNIT_CD, 4 as ORG_UNIT_LEVEL
from TBL
where cd_4 <> 0
union
select distinct cd_3 as ORG_UNIT_CD, upper('"'||nm_3||'"') as ORG_UNIT_NM, cd_2 as PARENT_ORG_UNIT_CD, 3 as ORG_UNIT_LEVEL
from TBL
where cd_3 <> 0
union
select distinct cd_2 as ORG_UNIT_CD, upper('"'||nm_2||'"') as ORG_UNIT_NM, cd_1 as PARENT_ORG_UNIT_CD, 2 as ORG_UNIT_LEVEL
from TBL
where cd_2 <> 0
union
select distinct cd_1 as ORG_UNIT_CD, upper('"'||nm_1||'"') as ORG_UNIT_NM, 0 as PARENT_ORG_UNIT_CD, 1 as ORG_UNIT_LEVEL
from TBL
where cd_1 <> 0
order by ORG_UNIT_LEVEL
我需要一个简化的查询而不是这个。 Becoz我只给了五级heirachy，实际上我有20级。
4 个答案:
答案 0 :(得分：4)
以下是适用于您的数据的UNPIVOT（适用于Oracle 11+版本）示例：
DROP TABLE mytable;
CREATE TABLE mytable(
OBJECT_ID INTEGER NOT NULL PRIMARY KEY
,CD_1 INTEGER NOT NULL
,NM_1 VARCHAR(15) NOT NULL
,CD_2 INTEGER NOT NULL
,NM_2 VARCHAR(28) NOT NULL
,CD_3 INTEGER NOT NULL
,NM_3 VARCHAR(20) NOT NULL
,CD_4 INTEGER NOT NULL
,NM_4 VARCHAR(28) NOT NULL
,CD_5 INTEGER NOT NULL
,NM_5 VARCHAR(16) NOT NULL
,DT VARCHAR(14) NOT NULL
);
insert into mytable
SELECT 53027429 AS OBJECT_ID,53001326 AS CD_1,'CEO & President' AS NM_1,53000330 AS CD_2,'Merck Manufacturing Division' AS NM_2,53011048 AS CD_3,'EMEA Operations' AS NM_3,54001626 AS CD_4,'HSC PLANT DIRECTION' AS NM_4,53027385 AS CD_5,'HSC NOT STERILE' AS NM_5,'4/24/2013 8:22' AS DT from dual
UNION ALL
SELECT 54010819,53001326,'CEO & President',54002711,'Merck Consumer Care',54007326,'MCC Emerging Markets',54010308,'Consumer Care - Asia Pacific',54010819,'MCC Singapore','8/23/201211:26' from dual
UNION ALL
SELECT 53025491,53001326,'CEO & President',53000330,'Merck Manufacturing Division',53011048,'EMEA Operations',54003518,'Rathdrum - Plant Manager',53020040,'Engineering & FM','5/22/2012 6:24' from dual
UNION ALL
SELECT 53023206,53001326,'CEO & President',53000321,'Corporate Finance',53000333,'MMD Finance',54001411,'Operations',54004358,'MMD Finance EMEA','5/31/2012 6:51' from dual
UNION ALL
SELECT 53021255,53001326,'CEO & President',53008586,'Global Human Health',54001286,'EUCan1',53013126,'Mid-Europe 1',53019000,'MER-1 Balkans','3/14/2013 7:06' from dual;
commit;
select distinct cd, nm,
case when parent=cd then 0 else parent end as parent, lvl
from (
select * from mytable
unpivot (
(cd, nm, parent) for lvl in ((CD_1, NM_1, CD_1) as 1,(CD_2, NM_2, CD_1) as 2,(CD_3, NM_3, CD_2) as 3,(CD_4, NM_4, CD_3) as 4, (CD_5, NM_5, CD_4) as 5)
)
)
order by lvl;
输出：
CD NM PARENT LVL
---------- ---------------------------- ---------- ----------
53001326 CEO & President 0 1
53000321 Corporate Finance 53001326 2
53008586 Global Human Health 53001326 2
54002711 Merck Consumer Care 53001326 2
53000330 Merck Manufacturing Division 53001326 2
53011048 EMEA Operations 53000330 3
54001286 EUCan1 53008586 3
54007326 MCC Emerging Markets 54002711 3
53000333 MMD Finance 53000321 3
54010308 Consumer Care - Asia Pacific 54007326 4
54001626 HSC PLANT DIRECTION 53011048 4
53013126 Mid-Europe 1 54001286 4
54001411 Operations 53000333 4
54003518 Rathdrum - Plant Manager 53011048 4
53020040 Engineering & FM 54003518 5
53027385 HSC NOT STERILE 54001626 5
54010819 MCC Singapore 54010308 5
53019000 MER-1 Balkans 53013126 5
54004358 MMD Finance EMEA 54001411 5
19 rows selected.
答案 1 :(得分：1)
在Oracle-11g中使用 Unpivot ：
执行列到行操作，并使用decode来派生其他列值。
试试这个：
select distinct /*product_code,*/quantity as ORG_UNIT_CD,
decode(product_code,'cd1','"'||nm1||'"','cd2','"'||nm2||'"','cd3','"'||nm3||'"','cd4','"'||nm4||'"','cd5','"'||nm5||'"') as ORG_UNIT_NM,
decode(product_code,'cd1','0','cd2','cd1','cd3','cd2','cd4','cd3','cd5','cd4') as PARENT_ORG_UNIT_CD
from unpivot_test2
UNPIVOT(quantity FOR product_code in (cd1 as 'cd1',cd2 as 'cd2',cd3 as 'cd3', cd4 as 'cd4', cd5 as 'cd5'))
where quantity <> 0;;
答案 2 :(得分：0)
我怀疑你是在追求以下内容：
with dummy as (select 1 id
from dual
connect by level <= 5)
select case when d.id = 1 then t.cd_1
when d.id = 2 then t.cd_2
when d.id = 3 then t.cd_3
when d.id = 4 then t.cd_4
when d.id = 5 then t.cd_5
end org_unit_cd,
case when d.id = 1 then t.nm_1
when d.id = 2 then t.nm_2
when d.id = 3 then t.nm_3
when d.id = 4 then t.nm_4
when d.id = 5 then t.nm_5
end org_unit_nm,
case when d.id = 1 then 0
when d.id = 2 then t.cd_1
when d.id = 3 then t.cd_2
when d.id = 4 then t.cd_3
when d.id = 5 then t.cd_4
end parent_org_unit_cd,
d.id org_unit_level
from (select distinct cd_1,
cd_2,
cd_3,
cd_4,
cd_5,
'"'||upper(nm_1)||'"' nm_1,
'"'||upper(nm_2)||'"' nm_2,
'"'||upper(nm_3)||'"' nm_3,
'"'||upper(nm_4)||'"' nm_4,
'"'||upper(nm_5)||'"' nm_5
from tbl) t
cross join dummy d
where case when d.id = 1 then t.cd_1
when d.id = 2 then t.cd_2
when d.id = 3 then t.cd_3
when d.id = 4 then t.cd_4
when d.id = 5 then t.cd_5
end != 0;
这简单地将每个不同的行从TBL复制5次，并在相关行上输出每个cd_N，nm_N和cd_n-1。以上将适用于10g和11g。
以下答案适用于11g：
select org_unit_cd,
org_unit_nm,
parent_org_unit_cd,
org_unit_level
from (select distinct cd_1,
cd_2,
cd_3,
cd_4,
cd_5,
'"'||upper(nm_1)||'"' nm_1,
'"'||upper(nm_2)||'"' nm_2,
'"'||upper(nm_3)||'"' nm_3,
'"'||upper(nm_4)||'"' nm_4,
'"'||upper(nm_5)||'"' nm_5,
0 cd_0
from tbl)
unpivot ((org_unit_cd, org_unit_nm, parent_org_unit_cd) for org_unit_level in ((cd_1, nm_1, cd_0) as 1,
(cd_2, nm_2, cd_1) as 2,
(cd_3, nm_3, cd_2) as 3,
(cd_4, nm_4, cd_3) as 4,
(cd_5, nm_5, cd_4) as 5))
答案 3 :(得分：0)
阿曼打败了我的白痴，所以我赞成他，但这里有一个例子，你可以跑去查看结果
WITH sample_data
AS (SELECT 'Smith' family_name,
'Pappy Smith' Great_Grandparent,
'Fred Smith' GrandParent,
'Joey Smith' Parent,
'Joey Jr' Child
FROM DUAL
UNION ALL
SELECT 'Jones',
'Great Grammy Jones',
'Grammy Jones',
'Bob',
'Bobby'
FROM DUAL)
SELECT family_name, source, val
FROM sample_data UNPIVOT INCLUDE NULLS (val
FOR (source)
IN (Great_Grandparent AS 'Great_Grandparent',
Grandparent AS 'Granparent',
Parent AS 'Parent',
Child AS 'Child') )
ORDER BY family_name,
DECODE (source,
'Great_Grandparent', 1,
'Granparent', 2,
'Parent', 3,
4)
相关问题
最新问题
- 我写了这段代码，但我无法理解我的错误
- 我无法从一个代码实例的列表中删除 None 值，但我可以在另一个实例中。为什么它适用于一个细分市场而不适用于另一个细分市场？
- 是否有可能使 loadstring 不可能等于打印？卢阿
- java中的random.expovariate()
- Appscript 通过会议在 Google 日历中发送电子邮件和创建活动
- 为什么我的 Onclick 箭头功能在 React 中不起作用？
- 在此代码中是否有使用“this”的替代方法？
- 在 SQL Server 和 PostgreSQL 上查询，我如何从第一个表获得第二个表的可视化
- 每千个数字得到
- 更新了城市边界 KML 文件的来源？