I'm new to SQL and after designing the database, i'm having trouble with some queries. The query i'm currently struggling with states:
"A list of the customers who have ordered at least one project with a higher than average expected duration."
SELECT Customer.name
FROM Project, Customer
WHERE Project.c_id = Customer.c_id AND Project.exp_duration > AVG(Project.exp_duration)
I tried to implement this code but i keep gettin the following error message : "An aggregate may not appear in the WHERE clause unless it is in a subquery contained in a HAVING clause or a select list, and the column being aggregated is an outer reference."
Can someone help me with this? I've thought about using joins but i can't get it to work either.
Thanks in advance!
Replace the table variables (#Project & #Customer) with your real tables (Project & Customer).
DECLARE #Project TABLE
(
p_id INT,
exp_duration DECIMAL(18,2),
c_id INT
)
DECLARE #Customer TABLE
(
c_id INT,
name VARCHAR(20)
)
INSERT #Project VALUES (1, 10, 1), (2, 5, 1), (3, 20, 1), (4, 10, 2), (5, 15, 2), (6, 20, 1)
INSERT #Customer VALUES (1, 'C1'), (2, 'C2')
-- average duration
-- SELECT AVG(exp_duration) FROM #Project
SELECT DISTINCT C.name
FROM #Customer C INNER JOIN #Project P ON C.c_id = P.c_id
WHERE p.exp_duration > (SELECT AVG(exp_duration) FROM #Project)
The following query gives the list of Customers who have ordered at least one project (i.e. being a part of one or more projects) and whose ExpectedDuration is greater than the Average ExpectedDuration.
I have used left outer join, group by, count and avg aggregate functions.
Select
C.CustomerID,
C.Name
From SampleCustomer C
Left Join SampleProject P
On C.CustomerID = P.CustomerID
Where P.ExpectedDuration > (Select Avg(ExpectedDuration) From SampleProject Where CustomerID = C.CustomerID)
Group By C.CustomerID, C.Name
Having Count(P.ProjectID) >= 1
Order By C.CustomerID;
Related
Consider I have two tables
Courses Program
---------------------------
course_ID program_id
course_title program_name
program_ID
Now, I want to check no of courses(by course_id) offered by each program (program_id).
select c.program_id ,p.program_name, count(course_id)
from courses c
join Program p on c.Program_id =p.Program_id
group by program_id,program_name
If I understood you correctly, you're searching for a GROUP BY and a corresponding aggregate.
--Creating sample tables and data
SELECT course_ID, course_title, program_ID
INTO #courses
FROM (
VALUES (0, 'course_0', 0),
(1, 'course_1', 0),
(2, 'course_2', 0),
(3, 'course_3', 0),
(4, 'course_4', 1),
(5, 'course_5', 1),
(NULL, 'course_6', 1)
) AS C (course_ID, course_title, program_ID)
SELECT program_ID, program_title
INTO #programs
FROM (
VALUES (0, 'program_0'),
(1, 'program_1')
) AS P (program_ID, program_title)
and after that execute the query
SELECT P.program_title, COUNT(C.course_ID) AS courses_amount
FROM #courses C
INNER JOIN #programs P ON C.program_ID = P.program_ID
GROUP BY P.program_ID, P.program_title
So you basically GROUP BY the value to which you to aggregate to and COUNT the 'course_id'.
COUNT(C.course_ID) only counts actual values and will ignore NULLs.
If you want to count the NULLs as well, just use COUNT(*).
EDIT: Forgot the result...
So it'll look like this:
program_title
courses_amount
program_0
4
program_1
2
I want to show the names of the customers along with the names of those who referred them. I thought I was on the right track but the result is messed up. I tried conditions ReferredBy NOT NULL in WHERE clause, ReferredBy NOT NULL in ON clause - no luck.
I would appreciate any help!
sqlfiddle
CREATE TABLE Customers(
Id int NOT NULL,
Name varchar(50) NOT NULL,
ReferredBy int REFERENCES Customers(Id),
PRIMARY KEY (Id)
);
INSERT INTO Customers VALUES
(11, 'Peter', 22),
(22, 'Ariel', NULL),
(33, 'Tom', 11);
My approach:
SELECT c.Id, c.Name, c.ReferredBy, n.Name as ReferredBy_name
FROM Customers c
LEFT JOIN Customers n
ON c.Id = n.ReferredBy
Desired Output:
I think the JOIN conditions have the tables inverted:
SELECT c.Id, c.Name, c.ReferredBy, n.Name as ReferredBy_name
FROM Customers c LEFT JOIN
Customers n
ON n.Id = c.ReferredBy ;
Suppose I have the following table
DROP TABLE IF EXISTS #toy_example
CREATE TABLE #toy_example
(
Id int,
Pet varchar(10)
);
INSERT INTO #toy
VALUES (1, 'dog'),
(1, 'cat'),
(1, 'emu'),
(2, 'cat'),
(2, 'turtle'),
(2, 'lizard'),
(3, 'dog'),
(4, 'elephant'),
(5, 'cat'),
(5, 'emu')
and I want to fetch all Ids that have certain pets (for example either cat or emu, so Ids 1, 2 and 5).
DROP TABLE IF EXISTS #Pets
CREATE TABLE #Pets
(
Animal varchar(10)
);
INSERT INTO #Pets
VALUES ('cat'),
('emu')
SELECT Id
FROM #toy_example
GROUP BY Id
HAVING COUNT(
CASE
WHEN Pet IN (SELECT Animal FROM #Pets)
THEN 1
END
) > 0
The above gives me the error Cannot perform an aggregate function on an expression containing an aggregate or a subquery. I have two questions:
Why is this an error? If I instead hard code the subquery in the HAVING clause, i.e. WHEN Pet IN ('cat','emu') then this works. Is there a reason why SQL server (I've checked with SQL server 2017 and 2008) does not allow this?
What would be a nice way to do this? Note that the above is just a toy example. The real problem has many possible "Pets", which I do not want to hard code. It would be nice if the suggested method could check for multiple other similar conditions too in a single query.
If I followed you correctly, you can just join and aggregate:
select t.id, count(*) nb_of_matches
from #toy_example t
inner join #pets p on p.animal = t.pet
group by t.id
The inner join eliminates records from #toy_example that have no match in #pets. Then, we aggregate by id and count how many recors remain in each group.
If you want to retain records that have no match in #pets and display them with a count of 0, then you can left join instead:
select t.id, count(*) nb_of_records, count(p.animal) nb_of_matches
from #toy_example t
left join #pets p on p.animal = t.pet
group by t.id
How about this approach?
SELECT e.Id
FROM #toy_example e JOIN
#pets p
ON e.pet = p.animal
GROUP BY e.Id
HAVING COUNT(DISTINCT e.pet) = (SELECT COUNT(*) FROM #pets);
I have a table called Customer, which has two columns, CustomerID and CustomerName. shown below
declare #customer table (CustomerID INT, CustomerName VARCHAR(10))
INSERT INTO #customer VALUES (1, 'Shane'), (2, 'Daniel'), (3, 'Karim'), (4, 'Eric'), (5, 'Zoe'), (6, 'Jack')
and another table called Age which has two columns, CustomerID and Age.
declare #age table (CustomerID INT, Age INT)
INSERT INTO #Age VALUES (1, 20), (2, 19), (3, 12), (6, 30)
Now, I want to add an additional column called HasAge on the Customer table, if CustomerID in Customer table is also exist in Age table, then return 1 else return 0.
I have tried to use left join and CTEs to get the result. So I am wondering if there is another way to do it (without using Joins). for example. use subquery in a select statement. Actually the real case for me is more complicated. for demonstration, I use the simpler example. My goal is not to use Joins and CTEs, all done in single query or other magic functions (maybe EXISTS ?).Thanks for any help !
The left join is concise, elegant, efficient:
select c.*, (case when a.customerid is null then 0 else 1 end) as has_age
from #customer c left join
#age a
on c.customerid = a.customerid;
That said, such a query is often written using case when exists, just in case more than one row matches:
select c.*
(case when exists (select 1 from #age a where c.customerid = a.customerid)
then 1 else 0
end) as has_age
from #customer c;
That too is elegant, concise, and efficient. If you know there is at most one row that matches, many people would prefer the left join version. Personally, I'm neutral.
When I say "efficient", I mean the performance is the best that can be expected. For real performance, you want an index on age(customerid). Both queries can make use of such an index.
I'm having some trouble with summing some column in my query with aggregation.
It's a bit difficult to describe what is happening but I'll try my best:
I have 3 tables - details, extra details and places.
Places is a table that contains places in the world. Details contains details about events that happened, and extra details provides some more data on the events. Each place has an ID and a ParentID (Like New York has an ID and it's parent ID is the US. Something like that). The ID of the event(details) appears a number of times as a column in the extra details table. The extra details table also holds the ID of the place that that event occurred at.
OK after all of that, what I'm trying to achieve is, for each place, the sum of the events that happened there. I know it sounds very specific, but it's what the client asked. Anyhow, example of what I'm trying to get to:
NewYork 60, Chicago 20, Houston 10 Then the US will have 90. And it has several levels.
So this is what I was trying to do:
With C(ID, NAME, COUNTT, ROOT_ID) as
(
SELECT d.ID, d.NAME,
(SELECT COUNT(LX.ID) as COUNTT
FROM EXTRA LX
RIGHT JOIN d ON LX.PLACE_ID = d.ID -- ****
GROUP BY d.ID, d.NAME),
d.ID as ROOT_ID
FROM PLACES d
UNION ALL
SELECT d.ID, d.NAME,
(SELECT COUNT(LX.ID) as COUNTT
FROM EXTRA LX
RIGHT JOIN d ON LX.PLACE_ID = d.ID
GROUP BY d.ID, d.NAME),
C.ROOT_ID
FROM PLACES dx
INNER JOIN C ON dx.PARENT_ID = C.ID
)
SELECT p.ID, p.NAME, S.SumIncludingChildren
FROM places p
INNER JOIN (
SELECT ROOT_ID, SUM(COUNTT) as SumIncludingChildren
FROM C
GROUP BY ROOT_ID
) S
ON p.ID = S.ROOT_ID
ORDER BY p.ID;
The details table is only for showing their data. I'll add that later. It's only comparing the respective columns. To making it work I don't need that. Only for the site data.
It doesn't work because it doesn't recognizes the 'd' where the '****' is. If I'll put a 'new instance' of that table, it won't work either. So I tried to replicate what the right join by doing 'NOT EXISTS IN' on a query that gets all the places instead of the right join...on. Same problem.
Maybe I don't get something. But I'm really seeking a solution and some explanation. I know my code isn't perfect.
Thanks in advance.
EDIT: I'm using OracleSQL on Toad 10.6
create table p(id number, up number, name varchar2(100));
create table e(id number, pid number, dsc varchar2(100));
insert into p values (1, null, 'country');
insert into p values (2, 1, 'center');
insert into p values (3, 1, 'province');
insert into p values (4, 2, 'capital');
insert into p values (5, 2, 'suburb');
insert into p values (6, 3, 'forest');
insert into p values (7, 3, 'village');
insert into p values (8, 7, 'shed');
insert into p values (9, 2, 'gov');
insert into e values (1, 8, 'moo');
insert into e values (2, 8, 'clank');
insert into e values (3, 7, 'sowing');
insert into e values (4, 6, 'shot');
insert into e values (5, 6, 'felling');
insert into e values (6, 5, 'train');
insert into e values (7, 5, 'cottage');
insert into e values (8, 5, 'rest');
insert into e values (9, 4, 'president');
insert into e values (10,1, 'GNP');
commit;
with
places as
(select id,
up,
connect_by_root id as root,
level lvl
from p
connect by prior id = up),
ev_stats as
(select root as place, max(lvl) as max_lvl, count(e.id) as ev_count
from places left outer join e
on places.id = e.pid
group by root)
select max_lvl, p.name, ev_count
from ev_stats inner join p on p.id = ev_stats.place
order by max_lvl desc;