Reputation: 2914
I have a very basic LEFT OUTER JOIN
to return all results from the left table and some additional information from a much bigger table. The left table contains 4935 records yet when I LEFT OUTER JOIN it to an additional table the record count is significantly larger.
As far as I'm aware it is absolute gospel that a LEFT OUTER JOIN
will return all records from the left table with matched records from the right table and null values for any rows which cannot be matched, as such it's my understanding that it should be impossible to return more rows than exist in the left table, but it's happening all the same!
SQL query is as follows:
SELECT
SUSP.Susp_Visits.SuspReason, SUSP.Susp_Visits.SiteID
FROM
SUSP.Susp_Visits
LEFT OUTER JOIN
DATA.Dim_Member ON SUSP.Susp_Visits.MemID = DATA.Dim_Member.MembershipNum
Perhaps I have made a mistake in the syntax or my understanding of LEFT OUTER JOIN
is incomplete, hopefully someone can explain how this could be occurring?
Upvotes: 270
Views: 384466
Reputation: 102538
A LEFT OUTER JOIN will return all records from the LEFT table joined with the RIGHT table where possible.
If there are matches, though, it will still return all rows that match. Therefore, one row in the LEFT table that matches two rows in the RIGHT table will return as two rows, just like an INNER JOIN.
Looking at the edited query for this specific question, it appears you are only returning data from the LEFT table. Therefore, if you only want data from the LEFT table, and you only want one row returned for each row in the LEFT table, then you have no need to perform a JOIN at all and can just do a SELECT directly from the LEFT table.
Upvotes: 276
Reputation: 1517
Simply, LEFT OUTER JOIN is the Cartesian product within each join key, along with the unmatched rows of the left table (i.e. for each key_x that has N records in table_L and M records in table_R the result will have N*M records if M>0, or N records if M=0)
Upvotes: 0
Reputation: 5256
Table1 Table2
_______ _________
1 2
2 2
3 5
4 6
SELECT Table1.Id,
Table2.Id
FROM Table1
LEFT OUTER JOIN Table2 ON Table1.Id=Table2.Id
Results:
1,null
2,2
2,2
3,null
4,null
Upvotes: 201
Reputation: 5908
The only way your query would return more number of rows than the left table ( which is SUSP.Susp_Visits
in your case), is that the condition (SUSP.Susp_Visits.MemID = DATA.Dim_Member.MembershipNum
) is matching multiple rows in the right table, which is DATA.Dim_Member
. So, there are multiple rows in the DATA.Dim_Member
where identical values are present for DATA.Dim_Member.MembershipNum
. You can verify this by executing the below query:
select DATA.Dim_Member.MembershipNum, count(DATA.Dim_Member.MembershipNum) from DATA.Dim_Member group by DATA.Dim_Member.MembershipNum
Upvotes: 1
Reputation: 453
Since the left table contains 4935 records, I suspect you want your results to return 4935 records. Try this:
create table table1
(siteID int,
SuspReason int)
create table table2
(siteID int,
SuspReason int)
insert into table1(siteID, SuspReason) values
(1, 678),
(1, 186),
(1, 723)
insert into table2(siteID, SuspReason) values
(1, 678),
(1, 965)
select distinct t1.siteID, t1.SuspReason
from table1 t1 left join table2 t2 on t1.siteID = t2.siteID and t1.SuspReason = t2.SuspReason
union
select distinct t2.siteID, t2.SuspReason
from table1 t1 right join table2 t2 on t1.siteID = t2.siteID and t1.SuspReason = t2.SuspReason
Upvotes: 0
Reputation: 96630
It isn't impossible. The number of records in the left table is the minimum number of records it will return. If the right table has two records that match to one record in the left table, it will return two records.
Upvotes: 65
Reputation: 17090
If you need just any one row from the right side
SELECT SuspReason, SiteID FROM(
SELECT SUSP.Susp_Visits.SuspReason, SUSP.Susp_Visits.SiteID, ROW_NUMBER()
OVER(PARTITION BY SUSP.Susp_Visits.SiteID) AS rn
FROM SUSP.Susp_Visits
LEFT OUTER JOIN DATA.Dim_Member ON SUSP.Susp_Visits.MemID = DATA.Dim_Member.MembershipNum
) AS t
WHERE rn=1
or just
SELECT SUSP.Susp_Visits.SuspReason, SUSP.Susp_Visits.SiteID
FROM SUSP.Susp_Visits WHERE EXISTS(
SELECT DATA.Dim_Member WHERE SUSP.Susp_Visits.MemID = DATA.Dim_Member.MembershipNum
)
Upvotes: 10
Reputation: 59
Pay attention if you have a where clause on the "right side' table of a query containing a left outer join... In case you have no record on the right side satisfying the where clause, then the corresponding record of the 'left side' table will not appear in the result of your query....
Upvotes: 5
Reputation: 4657
In response to your postscript, that depends on what you would like.
You are getting (possible) multiple rows for each row in your left table because there are multiple matches for the join condition. If you want your total results to have the same number of rows as there is in the left part of the query you need to make sure your join conditions cause a 1-to-1 match.
Alternatively, depending on what you actually want you can use aggregate functions (if for example you just want a string from the right part you could generate a column that is a comma delimited string of the right side results for that left row.
If you are only looking at 1 or 2 columns from the outer join you might consider using a scalar subquery since you will be guaranteed 1 result.
Upvotes: 21
Reputation: 19813
LEFT OUTER JOIN just like INNER JOIN (normal join) will return as many results for each row in left table as many matches it finds in the right table. Hence you can have a lot of results - up to N x M, where N is number of rows in left table and M is number of rows in right table.
It's the minimum number of results is always guaranteed in LEFT OUTER JOIN to be at least N.
Upvotes: 8
Reputation: 29153
if multiple (x) rows in Dim_Member are associated with a single row in Susp_Visits, there will be x rows in the resul set.
Upvotes: 2
Reputation: 882481
Each record from the left table will be returned as many times as there are matching records on the right table -- at least 1, but could easily be more than 1.
Upvotes: 15
Reputation: 156025
It seems as though there are multiple rows in the DATA.Dim_Member table per SUSP.Susp_Visits row.
Upvotes: 2
Reputation: 3558
Could it be a one to many relationship between the left and right tables?
Upvotes: 9