Opposite of inner join

What will be the opposite of inner join?

An OUTER join, which can be of three options:

• LEFT
• RIGHT
• FULL

This is a good visual representation of JOINs

I want to know the rows in Person with bad AddrId which don't have a row in the Address table.

Using LEFT JOIN/IS NULL

   SELECT p.*
     FROM PERSON p
LEFT JOIN ADDRESS a ON a.addrid = p.addrid
    WHERE a.addrid IS NULL

Using NOT EXISTS

SELECT p.*
  FROM PERSON p
 WHERE NOT EXISTS(SELECT NULL
                    FROM ADDRESS a
                   WHERE a.addrid = p.addrid)

Using NOT IN

SELECT p.*
  FROM PERSON p
 WHERE p.addrid NOT IN (SELECT a.addrid
                          FROM ADDRESS a)

If you consider an inner join as the rows of two tables that meet a certain condition, then the opposite would be the rows in either table that don't.

For example the following would select all people with addresses in the address table:

SELECT p.PersonName, a.Address
FROM people p
JOIN addresses a
    ON p.addressId = a.addressId

I imagine the "opposite" of this would be to select all of the people without addresses and all addresses without people. However this doesn't seem to be what you are asking, you seem to be interested in just one component of this: all the people without an address in the addresses table.

For this a left join would be best:

SELECT p.PersonName
FROM people p
LEFT JOIN addresses a
   ON p.addressId = a.addressId
WHERE a.addressId IS NULL

Note that often some prefer to write it differently as in their opinion it is more readable (however in my experience with large tables it performs worse than the above way):

SELECT PersonName
FROM people
WHERE addressId NOT IN (SELECT addressId FROM addresses)

An Inner join is not diametric to an Outer Join. They serve different purposes. However, a common pattern to find rows from one table that do not exist in another is to use an Outer Join:

Select ...
From Table1
    Left Join Table2
        On Table2.ForeignKeyCol = Table1.PrimaryKeyCol
Where Table2.PrimaryKeyCol Is Null

This returns all rows from Table1 and any matching rows from Table2 such that if a given Table1 row has no Table2 match, a null for the Table2 columns are returned. By then requiring that a non-nullable column (Table2.PrimaryKeyCol) be Null, I will get all rows from Table1 that do not exist in Table2. Using your example table names we would have something like:

Select ...
From Person
    Left Join Address
        On Address.PersonId = Person.Id
Where Address.Id Is Null

I think the best solution would be using EXISTS. Like this:

SELECT * FROM Person P
WHERE P.AddrId IS NOT NULL 
  AND NOT EXISTS ( SELECT 1 FROM Address A WHERE A.AddrId = P.AddrId )

The query above will return every person that the AddrId is set but does not have a corresponding record in the Address table.

Obs.: Use the constant 1 in the EXISTS query to avoid table access.