Reputation: 479
I have a two separate tables, one with vacancies, and one with applications to those vacancies. I want to select a new table which selects from the vacancies table with a number of other columns from that table, and another column that calculates how many applications there are for those vacancies. So my vacancy table looks like this:
ID Active StartDate JobID JobTypeID HoursPerWeek
1 1 2017-02-28 2 CE 0
2 1 2017-02-15 4 CE 40
3 1 2017-02-14 1 CE 40
4 1 2017-02-28 1 CE 48
My applications table looks like this:
ID VacancyID Forename Surname EmailAddress TelephoneNumber
1 1 John Smith [email protected] 447777777777
2 2 John Smith [email protected] 447748772641
3 2 John Smith [email protected] 447777777777
4 2 John Smith [email protected] 447700123456
5 4 John Smith [email protected] 447400123569
6 4 John Smith [email protected] 447400126547
7 4 John Smith [email protected] 447555123654
I want a table that looks like this:
ID Active StartDate JobID HoursPerWeek NumberOfApplicants
1 1 2017-02-28 2 0 1
2 1 2017-02-15 4 40 3
3 1 2017-02-14 1 40 0
4 1 2017-02-28 1 48 3
How can I select that table using joins and count the number of applicants where the VacancyID is equal to the ID of the first vacancy table? I have tried:
select Vacancy.ID, VacancyID, count(*) as NumberOfApplications from VacancyApplication
join Vacancy on Vacancy.ID=VacancyID
group by VacancyID, Vacancy.ID
This obviously doesn't select all the other columns and it also does not select ID 3 because there are 0 applications for that - I want ID 3 to be there with a value of 0 as well as all the other columns. How do I do this? I've tried various forms of grouping and selecting but I'm quite new to SQL so I'm not really sure how this can be done.
Upvotes: 0
Views: 58
Reputation: 1270391
Use left join
and table aliases:
select v.ID, count(va.VacancyID) as NumberOfApplications
from Vacancy v join
VacancyApplication va
on v.ID = va.VacancyID
group by v.ID;
You seem to want all the columns. You could include them in the group by
. However, a correlated subquery or outer apply
is simpler:
select v.*, va.cnt
from vacancy v outer apply
(select count(*) as cnt
from VacancyApplication va
where v.ID = va.VacancyID
) va;
This is probably more efficient anyway, especially if you have an index on VacancyApplication(VacancyID)
.
Upvotes: 0
Reputation: 82504
You can use count as a window function using the OVER
clause, thus eliminating he need for group by:
SELECT v.ID,
v.Active,
v.StartDate,
v.JobID,
v.JobTypeID,
COUNT(va.ID) OVER(PARTITION BY v.ID) HoursPerWeek
FROM Vacancy v
LEFT JOIN vacancyapplication va ON(v.ID = va.VacancyID)
Upvotes: 0
Reputation: 2348
Hoping, i understood your problem correctly. Please try below query
select Vacancy.ID, VacancyID, count(*) as NumberOfApplications from VacancyApplication
left join Vacancy on Vacancy.ID=VacancyID
group by VacancyID, Vacancy.ID
Upvotes: 0
Reputation: 93734
Use RIGHT JOIN
instead of INNER JOIN
and count the vacancyid
column from vacancyapplication
table. For the non matching records you will get count as 0
SELECT v.id, v.Active, v.StartDate, v.JobID, v.HoursPerWeek
Count(va.vacancyid) AS NumberOfApplications
FROM vacancyapplication va
RIGHT JOIN vacancy v
ON v.id = va.vacancyid
GROUP BY v.id, v.Active, v.StartDate, v.JobID, v.HoursPerWeek
Start using Alias
names, it makes the query more readable
Upvotes: 2