I have a table that contains Home addresses and Mailing addresses. It looks like this:
ID Name StNum StName City State Zip Type
-- ---- ----- ------ ---- ----- --- ----
1 Joe 1234 Main St Waco TX 76767 HOM
1 Joe 2345 High St Waco TX 76763 MLG
2 Amy 3456 Broad St Athens GA 34622 HOM
3 Mel 987 Front St Cary NC 65331 HOM
3 Mel 1111 Main Ave Hilo HI 99779 MLG
I need to write an SQL statement that will only return the Mailing address (MLG record) if it exists, and if not, will return the Home address (HOM record).
The expected results from this table would be:
ID Name StNum StName City State Zip Type
-- ---- ----- ------ ---- ----- --- ----
1 Joe 2345 High St Waco TX 76763 MLG
2 Amy 3456 Broad St Athens GA 34622 HOM
3 Mel 1111 Main Ave Hilo HI 99779 MLG
Any help that you can provide would be much appreciated! Thanks!
use correlated subquery
select * from
(
select *,case when Type='MLG' then 1 else 0 end as typeval
from tablename
)A where typeval in (select max(case when Type='MLG' then 1 else 0 end) from tablename b
where a.name=b.name)
OR if your DB supports row_number()
then u can try below -
select * from
(
select *, row_number() over(partition by name order by case when Type='MLG' then 1 else 0 end desc)
from tablename
)A where rn=1
In case you are using SQL Server, i would solve it with the ROW_NUMBER function.
SELECT ID, Name, StNum, StName, City, State, Zip, Type
FROM (
SELECT *
,ROW_NUMBER() OVER (PARTITION BY ID ORDER BY Type DESC) AS Rn
FROM yourtable
)
WHERE Rn = 1
This can be done using a WHERE
clause that exclude the ids of the users who have MLG
Schema (MySQL v5.7)
CREATE TABLE test (
`ID` INTEGER,
`Name` VARCHAR(3),
`StNum` INTEGER,
`StName` VARCHAR(8),
`City` VARCHAR(6),
`State` VARCHAR(2),
`Zip` INTEGER,
`Type` VARCHAR(3)
);
INSERT INTO test
(`ID`, `Name`, `StNum`, `StName`, `City`, `State`, `Zip`, `Type`)
VALUES
('1', 'Joe', '1234', 'Main St', 'Waco', 'TX', '76767', 'HOM'),
('1', 'Joe', '2345', 'High St', 'Waco', 'TX', '76763', 'MLG'),
('2', 'Amy', '3456', 'Broad St', 'Athens', 'GA', '34622', 'HOM'),
('3', 'Mel', '987', 'Front St', 'Cary', 'NC', '65331', 'HOM'),
('3', 'Mel', '1111', 'Main Ave', 'Hilo', 'HI', '99779', 'MLG');
Query #1
SELECT id,
name,
StNum,
StName,
City,
State,
Zip,
Type
FROM test t1
WHERE t1.`Type` = 'MLG'
OR t1.id NOT IN
(
SELECT id
FROM test t2
WHERE t2.`Type` = 'MLG'
);
Output :
| id | name | StNum | StName | City | State | Zip | Type |
| --- | ---- | ----- | -------- | ------ | ----- | ----- | ---- |
| 1 | Joe | 2345 | High St | Waco | TX | 76763 | MLG |
| 2 | Amy | 3456 | Broad St | Athens | GA | 34622 | HOM |
| 3 | Mel | 1111 | Main Ave | Hilo | HI | 99779 | MLG |
Or, my first dumb version :
This can be done using UNION
Schema (MySQL v5.7)
CREATE TABLE test (
`ID` INTEGER,
`Name` VARCHAR(3),
`StNum` INTEGER,
`StName` VARCHAR(8),
`City` VARCHAR(6),
`State` VARCHAR(2),
`Zip` INTEGER,
`Type` VARCHAR(3)
);
INSERT INTO test
(`ID`, `Name`, `StNum`, `StName`, `City`, `State`, `Zip`, `Type`)
VALUES
('1', 'Joe', '1234', 'Main St', 'Waco', 'TX', '76767', 'HOM'),
('1', 'Joe', '2345', 'High St', 'Waco', 'TX', '76763', 'MLG'),
('2', 'Amy', '3456', 'Broad St', 'Athens', 'GA', '34622', 'HOM'),
('3', 'Mel', '987', 'Front St', 'Cary', 'NC', '65331', 'HOM'),
('3', 'Mel', '1111', 'Main Ave', 'Hilo', 'HI', '99779', 'MLG');
Query #1
SELECT id,
name,
StNum,
StName,
City,
State,
Zip,
Type
FROM test t1
WHERE t1.`Type` = 'MLG'
UNION ALL
SELECT id,
name,
StNum,
StName,
City,
State,
Zip,
Type
FROM test t2
WHERE t2.id NOT IN (SELECT id FROM test t3 WHERE t3.`Type` = 'MLG')
ORDER BY id;
Output
| id | name | StNum | StName | City | State | Zip | Type |
| --- | ---- | ----- | -------- | ------ | ----- | ----- | ---- |
| 1 | Joe | 2345 | High St | Waco | TX | 76763 | MLG |
| 2 | Amy | 3456 | Broad St | Athens | GA | 34622 | HOM |
| 3 | Mel | 1111 | Main Ave | Hilo | HI | 99779 | MLG |
This is a prioritization query. With two values, often the simplest method is union all
with not exists
(or not in
).
That does not generalize well For more values, using row_number()
with case
is convenient:
select t.*
from (select t.*,
row_number() over (partition by id
order by (case when type = 'MLG' then 1 else 2 end)
) as seqnum
from t
) t
where seqnum = 1;
In your particular case, you could use order by type desc
, because the two types happen to be prioritized in reverse alphabetical ordering. However, I recommend using case
because the intention is more explicit.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.