I am attempting to run this query
SELECT u.*, string_agg(CAST(uar.roleid AS VARCHAR(100)), ',') AS roleids, string_agg(CAST(r.role AS VARCHAR(100)), ',') AS systemroles
FROM idpro.users AS u
INNER JOIN idpro.userapplicationroles AS uar ON u.id = uar.userid
INNER JOIN idpro.roles AS r ON r.id = uar.roleid
GROUP BY u.id, uar.applicationid
HAVING u.organizationid = '77777777-f892-4f4a-8328-c31df32bd6ba'
AND uar.applicationid = 'd88fbf05-c048-4697-8bf3-036f39897183'
AND (u.statusid = '7f9f0b75-44b7-4216-bf2a-03abc47dcff8')
AND uar.roleid IN ('cc9ada1c-fa21-400b-be98-c563ebb65a9c','de087148-4788-43da-89e2-dd7dff097735');
However, I'm getting an error stating that
ERROR: column "uar.roleid" must appear in the GROUP BY clause or be used in an aggregate function
LINE 9: AND uar.roleid IN ('cc9ada1c-fa21-400b-be98-c563ebb65a9c','...
string_agg() IS an aggregate function, is it not? My intent, if it isn't obvious, is to return each user record with the roleids and rolenames in comma-delimited lists. If I am doing everything wrong, could you please point me in the right direction?
You are filtering the data, so a WHERE clause would be needed. This tutorial is worth reading.
SELECT u.*,
string_agg(CAST(uar.roleid AS VARCHAR(100)), ',') AS roleids,
string_agg(CAST(r.role AS VARCHAR(100)), ',') AS systemroles
FROM idpro.users AS u
INNER JOIN idpro.userapplicationroles AS uar ON u.id = uar.userid
INNER JOIN idpro.roles AS r ON r.id = uar.roleid
WHERE u.organizationid = '77777777-f892-4f4a-8328-c31df32bd6ba'
AND uar.applicationid = 'd88fbf05-c048-4697-8bf3-036f39897183'
AND (u.statusid = '7f9f0b75-44b7-4216-bf2a-03abc47dcff8')
AND uar.roleid IN ('cc9ada1c-fa21-400b-be98-c563ebb65a9c','de087148-4788-43da-89e2-dd7dff097735');
GROUP BY u.id, uar.applicationid
The HAVING clause is helpful for filtering the aggregated values or the groups.
Since you are grouping by u.id, the table primary key you have access to every column of the u table. You can either use a where clause or a having clause.
For uar.applicationid, it is part of the group by so you can also use either a where or a having.
uar.roleid is not part of the group by clause, so to be usable in the having clause, you would have to consider the aggregated value.
The following example filters out rows whose aggregated length is more than 10 chars.
HAVING length(string_agg(CAST(uar.roleid AS VARCHAR(100)), ',')) > 10
A more common usage, on numerical field, is to filter out if the number of aggregated rows is less than a threshold (having count(*) > 2) or a sum of some kind (having sum(vacation_days) > 21)
Related
There is an example request in which there are several joins.
SELECT DISTINCT ON(a.id_1) 1, a.name, b.task, c.created_at
FROM a
INNER JOIN b ON a.id_2 = b.id
INNER JOIN c ON a.ID_2 = c.id
WHERE a.deleted_at IS NULL
ORDER BY a.id_1 desc
In this case, the query will work, sorting by unique values of id_1 will take place. But I need to sort by the column a.name. In this case, postresql will swear with the words ERROR: SELECT DISTINCT ON expressions must match initial ORDER BY expressions.
The following query can serve as a solution to the problem:
SELECT *
FROM(
SELECT DISTINCT ON(a.id_1) a.name, b.task, c.created_at
FROM a
INNER JOIN b ON a.id_2 = b.id
INNER JOIN c ON a.ID_2 = c.id
WHERE a.deleted_at IS NULL
)
ORDER_BY a.name desc
But in reality the database is very large and such a query is not optimal. Are there other ways to sort by the selected column while keeping one uniqueness?
My query is below:
select
u.Id,
STRING_AGG(sf.Naziv, ', ') as 'Ustrojstvena jedinica',
ISNULL(CONVERT(varchar(200), (STRING_AGG(TRIM(p.Naziv), ', ')), 121), '')
as 'Partner',
from Ugovor as u
left join VezaUgovorPartner as vup
on vup.UgovorId = u.Id AND vup.IsDeleted = 'false'
left join [TEST_MaticniPodaci2].dbo.Partner as p
on p.PartnerID = vup.PartnerId
left join [dbo].[VezaUgovorUstrojstvenaJedinica] as vuu
on vuu.UgovorId = u.Id
left join [TEST_MaticniPodaci2].hcphs.SifZavod as sf
on sf.Id = vuu.UstrojstvenaJedinicaId
left join [dbo].[SifVrstaUgovora] as vu
on u.VrstaUgovoraId = vu.Id
group by u.Id, sf.Naziv
My problem is that I can have more sf.Naziv and also only one sf.Naziv so I have to check if there is one and then show only one result and if there is two or more to show more results. But for now the problem is when I have only one sf.Naziv, query returns two sf.Naziv with the same name because in first STRING_AGG i have more records about p.Naziv.
I have no idea how to implement DISTINCT into STRING_AGG function
Any other solutions are welcome, but I think it should work with DISTINCT function.
It looks like distinct won't work, so what you should do is put your whole query in a subquery, remove the duplicates there, then do STRING_AGG on the data that has no duplicates.
SELECT STRING_AGG(data)
FROM (
SELECT DISTINCT FROM ...
)
I like this format for distinct values:
(d is required but you can use any variable name there)
SELECT STRING_AGG(LoadNumber, ',') as LoadNumbers FROM (SELECT DISTINCT LoadNumber FROM [ASN]) d
A sample query to remove duplicates while using STRING_AGG().
WITH cte AS (
SELECT DISTINCT product
FROM activities
)
SELECT STRING_AGG(product, ',') products
FROM cte;
Or you can use the following query. The result is same -
SELECT STRING_AGG(product, ',') as products
from (
SELECT product
FROM Activities
GROUP BY product
) as _ ;
I have this SQL:
select * from `posts`
where `posts`.`deleted_at` is null
and `expire_at` >= '2017-03-26 21:23:42.000000'
and (
select count(distinct tags.id) from `tags`
inner join `post_tag` on `tags`.`id` = `post_tag`.`tag_id`
where `post_tag`.`post_id` = `posts`.`id`
and (`tags`.`tag` like 'PHP' or `tags`.`tag` like 'pop' or `tags`.`tag` like 'UI')
) >= 1
Is it possible order the results by number of tags in posts?
Maybe add there alias?
Any information can help me.
Convert your correlated subquery into a join:
select p.*
from posts p
join (
select pt.post_id,
count(distinct t.id) as tag_count
from tags t
inner join post_tag pt on t.id = pt.tag_id
where t.tag in ('PHP', 'pop', 'UI')
group by pt.post_id
) pt on p.id = pt.post_id
where p.deleted_at is null
and p.expire_at >= '2017-03-26 21:23:42.000000'
order by pt.tag_count desc;
Also, note that I changed the bunch of like and or to single IN because you are not matching any pattern i.e. there is no % in the string. So, better using single IN instead.
Also, if you have defined your table names, column names etc keeping keywords etc in mind, you shouldn't have the need to use the backticks. They make reading a query difficult.
I'm trying to get lateral to work in a Postgres 9.5.3 query.
select b_ci."IdOwner",
ci."MinimumPlaces",
ci."MaximumPlaces",
(select count(*) from "LNK_Stu_CI" lnk
where lnk."FK_CourseInstanceId" = b_ci."Id") as "EnrolledStudents",
from "Course" c
join "DBObjectBases" b_c on c."Id" = b_c."Id"
join "DBObjectBases" b_ci on b_ci."IdOwner" = b_c."Id"
join "CourseInstance" ci on ci."Id" = b_ci."Id",
lateral (select ci."MaximumPlaces" - "EnrolledStudents") x
I want the right-most column to be the result of "MaximumPlaces" - "EnrolledStudents" for that row but am struggling to get it to work. At the moment PG is complaining that "EnrolledStudents" does not exist - which is exactly the point of "lateral", isn't it?
select b_ci."IdOwner",
ci."MinimumPlaces",
ci."MaximumPlaces",
(select count(*) from "LNK_Stu_CI" lnk
where lnk."FK_CourseInstanceId" = b_ci."Id") as "EnrolledStudents",
lateral (select "MaximumPlaces" - "EnrolledStudents") as "x"
from "Course" c
join "DBObjectBases" b_c on c."Id" = b_c."Id"
join "DBObjectBases" b_ci on b_ci."IdOwner" = b_c."Id"
join "CourseInstance" ci on ci."Id" = b_ci."Id"
If I try inlining the lateral clause (shown above) in the select it gets upset too and gives me a syntax error - so where does it go?
Thanks,
Adam.
You are missing the point with LATERAL. It can access columns in tables in the FROM clause, but not aliases defined in SELECT clause.
If you want to access alias defined in SELECT clause, you need to add another query level, either using a subquery in FROM clause (AKA derived table) or using a CTE (Common Table Expression). As CTE in PostgreSQL acts as an optimization fence, I strongly recommend going with subquery in this case, like:
select
-- get all columns on the inner query
t.*,
-- get your new expression based on the ones defined in the inner query
t."MaximumPlaces" - t."EnrolledStudents" AS new_alias
from (
select b_ci."IdOwner",
ci."MinimumPlaces",
ci."MaximumPlaces",
(select count(*) from "LNK_Stu_CI" lnk
where lnk."FK_CourseInstanceId" = b_ci."Id") as "EnrolledStudents",
from "Course" c
join "DBObjectBases" b_c on c."Id" = b_c."Id"
join "DBObjectBases" b_ci on b_ci."IdOwner" = b_c."Id"
join "CourseInstance" ci on ci."Id" = b_ci."Id"
) t
I have the following subquery.
In the end i want to count for every agent for all this users the controllers.
So in the substring i count the controllers for all the agent his users and
then i want to sum them up.
The substring itself is running, however i get this error when i run the complete query
ERROR: missing FROM-clause entry for table "sub"
LINE 2: a.id as a_id, SUM(sub.c_count ) as c_count,
SELECT
a.id as a_id, SUM(sub.c_count ) AS c_count,
(
SELECT u.id, COUNT(c.user_id) AS c_amount
FROM users u
JOIN controllers c ON (c.user_id = u.id)
GROUP BY
u.id
)sub
from agents a;
You are getting this error because the subquery which you have aliased as sub returns not only multiple records, but also multiple values per record, but it appears in the SELECT list. Generally, only scalars (single values) or subqueries which return a single value can appear in the SELECT list. You have two options, you can either modify sub to return a single value, or you can try to join this subauery to the outer query. Given that you mentioned the subquery is supposed to do some aggregation for each user or agent, my guess is that joining is the answer. In the query below I join sub to the outer query on the agent ID matching the user ID from sub.
SELECT a.id AS a_id,
COALESCE(sub.c_amount, 0) AS c_count
FROM agents a
LEFT JOIN
(
SELECT u.id, COUNT(c.user_id) AS c_amount
FROM users u
INNER JOIN controllers c
ON c.user_id = u.id
GROUP BY u.id
) sub
ON a.id = sub.id