sql join if value exists in other table then Count it - tsql

I have following tables.
Table A
UserID | key 1 | A 2 | B 3 | A 4 | C 5 |
Table B
UserID | Num1 | 501 | 3002 |3 | 1004 | 20
I have query like this
SELECT COUNT(key) AS cnt, key
FROM A
WHERE key <> ''
GROUP BY key
ORDER BY cnt DESC
The results should be something like this
key | cnt A | 2 B | 1 C | 1
What I would like to add is Joining Table B.
If UserID has value in Num in Table B, I would like to count UserID with/Num Grouped by key
Here is desired results
key | cnt | Has Num? A | 2 | 2 B | 1 | 0 C | 1 | 1
I tried to write subquery but I can't attach it to main query. Subquery is something like this.
SELECT COUNT(DISTINCT UserID) AS num
FROM B
LEFT OUTER JOIN A ON B.UserID = A.UserID
WHERE Num <>'' AND key <> ''
GROUP BY key


If I'm understanding this correctly, what you're looking for is a count of the Keys in Table A when they were used by a UserID, and then a count of the number of unique UserIDs in Table B who both appeared in the first Table A query and had a Num.
Try this:
SELECT a.[Key], COUNT(a.[Key]) AS cnt, isNull(SUM(b.bCnt), 0) AS [Has Num?]
FROM #TableA a
LEFT OUTER JOIN (
SELECT b.UserID, 1
FROM #TableB b
WHERE LEN(b.Num) > 0
GROUP BY b.UserID
) b (UserID, bCnt) ON b.UserID = a.UserID
WHERE LEN(a.[Key]) > 0
GROUP BY a.[Key]
This query gives the results that you were expecting.


DECLARE #TableA TABLE(UserID INT, [Key] CHAR(1))
INSERT INTO #TableA VALUES(1,'A'),(2,'B'),(3,'A'),(4,'C'),(5,'')
DECLARE #TableB TABLE(UserID INT, Num INT NULL)
INSERT INTO #TableB VALUES(1,50),(1,300),(2,NULL),(3,100),(4,20)
SELECT x.[Key],x.Cnt,y.[Has Num?]
FROM
( SELECT [Key],Cnt = COUNT([Key])
FROM #TableA
WHERE LEN([Key])>0
GROUP BY [Key]
)X
JOIN
(
SELECT a.[Key],[Has Num?] = COUNT(b.Num)
FROM #TableA a
JOIN #TableB b ON a.UserID = b.UserID
GROUP BY a.[Key]
)Y
ON x.[Key] = Y.[Key]
Key Cnt Has Num?
A 2 3
B 1 0
C 1 1


How about an OUTER APPLY
SELECT [Key], COUNT(a.[Key]) AS cnt, SUM(x.NumCount) AS [Has Num?]
FROM #TableA a
OUTER APPLY (SELECT COUNT(NUM) AS NumCount
FROM #TableB b
WHERE b.UserId = a.UserId AND Num IS NOT NULL
) x
WHERE [Key] <> ''
GROUP BY [Key]
ORDER BY cnt DESC
Result:
Key cnt Has Num?
---- ----------- -----------
A 2 3
B 1 0
C 1 1

Related

removing duplicate rows and dependencies without cursor

I have a table that has a long list of duplicated items. I am working on a stored procedure to consolidate them all into one record. Each one of the duplicated items has a number of child tables that should either be deleted, or rekeyed to point at the resulting record. My table has an Id, but the ReadableIdentifier is the column I need to deduplicate.
Id | ReadableIdentifier | Name | UpdatedOn
1 | ABC1234 | Product X | 2014-04-25 16:00:08.000
2 | ABC1234 | Product X | 2014-04-28 16:00:08.000
3 | ABC1234 | Product X | 2014-04-21 16:00:08.000
4 | ABDD9945 | Widget R | 2014-04-25 16:00:08.000
5 | ABDD9945 | Widget R | 2014-04-25 18:45:08.000
As you can see, records 1-3 are duplicates with different Id's and UpdatedOn dates. Same for 4-5. I need to consolidate these into one record, preferring the one with the most recent UpdatedOn date.
End Goal (not showing children tables):
Id | ReadableIdentifier | Name | UpdatedOn
2 | ABC1234 | Product X | 2014-04-28 16:00:08.000
5 | ABDD9945 | Widget R | 2014-04-25 18:45:08.000
I am using a CURSOR to do this, but am wondering if there is a better solution.
DECLARE dupeCursor CURSOR
FAST_FORWARD
FOR
WITH Counts AS (
SELECT
COUNT(1) Count,
ReadableIdentifier
FROM dbo.Item WITH (NOLOCK)
WHERE ReadableIdentifier IS NOT NULL
GROUP BY ReadableIdentifier)
SELECT
Counts.Count,
Counts.ReadableIdentifier,
Counts.CompanyId
FROM
Counts
WHERE Counts.Count > 1;
OPEN dupeCursor;
DECLARE #readableId VARCHAR(50);
DECLARE #itemToPersistId INT, #itemToDeleteId INT;
FETCH NEXT FROM dupeCursor INTO #readableId;
WHILE ##FETCH_STATUS = 0
BEGIN
WITH V AS (
SELECT Id, ROW_NUMBER() OVER (PARTITION BY ReadableId ORDER BY UpdatedOn DESC) as Row
FROM dbo.Item WITH (NOLOCK) WHERE ReadableId = #readableId
)
SELECT #itemToPersistId = Id
FROM V
WHERE V.Row = 1
CREATE TABLE #itemsToDelete (Id UNIQUEIDENTIFIER)
INSERT INTO #itemsToDelete
SELECT Id
FROM dbo.Item WITH (NOLOCK)
WHERE ReadableId = #readableId AND Id != #itemToPersistId;
--UPDATE CHILDREN TABLES
DELETE FROM dbo.ItemDetails WHERE ItemId IN (SELECT Id FROM #itemsToDelete);
UPDATE dbo.ItemPurchases SET ItemId = #itemToPersistId
WHERE ItemId IN (SELECT Id FROM #itemsToDelete);
UPDATE dbo.PurchaseOrders SET ItemId = #itemToPersistId
WHERE ItemId IN (SELECT Id FROM #itemsToDelete);
DELETE FROM dbo.ItemMetadata WHERE ItemId IN (SELECT Id FROM #itemsToDelete);
--delete Duplicated Items
DELETE FROM dbo.Item WHERE Id IN (SELECT Id FROM #itemsToDelete);
DROP TABLE #itemsToDelete
FETCH NEXT FROM dupeCursor INTO #readableId;
END
CLOSE dupeCursor;
DEALLOCATE dupeCursor;
I realize the cursor is most likely the issue, but I'm not sure how to go about updating all of the child tables without using one.

Ok I dont have data to test this for the child tables but it should work:
WITH V
AS (SELECT *,
ROW_NUMBER() OVER(PARTITION BY ReadableId ORDER BY UpdatedOn DESC) AS Row
FROM dbo.Item WITH (NOLOCK))
SELECT *
INTO #itemsToDelete
FROM V;
--UPDATE CHILDREN TABLES
DELETE FROM dbo.ItemDetails
WHERE ItemId IN
(
SELECT Id
FROM #itemsToDelete
WHERE Row > 1
);
UPDATE IP
SET
IP.ItemId = itk.ID
FROM dbo.ItemPurchases AS IP
INNER JOIN #itemsToDelete AS itd ON IP.ItemId = itd.ID
AND itd.Row > 1
INNER JOIN #itemsToDelete AS itk ON itk.ReadableIdentifier = itd.ReadableIdentifier
AND itk.Row = 1
AND itd.Row > 1;
UPDATE po
SET
po.ItemId = itk.ID
FROM dbo.PurchaseOrders AS po
INNER JOIN #itemsToDelete AS itd ON po.ItemId = itd.ID
AND itd.Row > 1
INNER JOIN #itemsToDelete AS itk ON itk.ReadableIdentifier = itd.ReadableIdentifier
AND itk.Row = 1
AND itd.Row > 1;
DELETE FROM dbo.ItemMetadata
WHERE ItemId IN
(
SELECT Id
FROM #itemsToDelete
WHERE Row > 1
);
--delete Duplicated Items
DELETE FROM dbo.Item
WHERE Id IN
(
SELECT Id
FROM #itemsToDelete
WHERE Row > 1
);

Capture First Character of Last Group of 1s in a Binary Series Part II: Multiple IDs

I have data something like this:
ID 1 1 1 1 1 1 1 1 1 1 1 1
Month J F M A M J J A S O N D
Status 1 0 0 1 0 1 0 0 1 1 1 1
ID 2 2 2 2 2 2 2 2 2 2 2 2
Month J F M A M J J A S O N D
Status 1 0 1 0 1 0 1 0 1 0 1 1
ID 3 3 3 3 3 3 3 3 3 3 3 3
Month J F M A M J J A S O N D
Status 0 0 0 0 0 0 0 0 0 0 0 1
Using t-SQL, I am trying to capture the month corresponding to the first STATUS = 1 in the last group of 1s for each ID, i.e., September, November and December in this example.
Here is the code I'm using:
IF OBJECT_ID('tempdb..#Temp1') IS NOT NULL DROP TABLE #Temp1
;WITH PARTITIONED1 AS
(SELECT t0.ID
, t0.Year_Month
, LAST_VALUE(t0.Year_Month) OVER (PARTITION BY t0.Account_Number ORDER BY t0.Year_Month) AS STATUS
, ROW_NUMBER() OVER (PARTITION BY t0.Account_Number ORDER BY t0.Year_Month) AS rn1
FROM #Temp0 t0
)
SELECT *
INTO #Temp1
FROM PARTITIONED1 p1
ORDER BY t0.ID
, t0.Year_Month
IF OBJECT_ID('tempdb..#Temp') IS NOT NULL DROP TABLE #Temp
SELECT *
INTO #Temp
FROM #Temp1 t1
WHERE t1.rn1 = (SELECT MAX(b.rn1) + 1 FROM #Temp1 b WHERE b.STATUS = 0)
GROUP BY t1.ID
, t1.Year_Month
, t1.rn1
However, this just returns the last instance where STATUS = 1 is achieved overall as the first 1 of the last group of 1s, in this case January.
I've tried using CASE statements and grouping in various combinations (hence the intermediate step reading the data into #Temp1), but have not been able to get results for all three IDs; is anyone able to assist?
Thanks in advance!

Assuming Ju for June and Jl for July:
--Sample Data
IF OBJECT_ID('tempdb..#Temp0') IS NOT NULL DROP TABLE #Temp0
CREATE TABLE #Temp0 (ID INT, Year_Month VARCHAR(1), Status INT)
INSERT INTO #Temp0
VALUES(1,'J',1),(1,'F',0),(1,'M',0),(1,'A',1),(1,'M',0),(1,'J',1),(1,'J',0),(1,'A',0),(1,'S',1),(1,'O',1),(1,'N',1),(1,'D',1),(2,'J',1),(2,'F',0),(2,'M',1),(2,'A',0),(2,'M',1),(2,'J',0),(2,'J',1),(2,'A',0),(2,'S',1),(2,'O',0),(2,'N',1),(2,'D',1),(3,'J',0),(3,'F',0),(3,'M',0),(3,'A',0),(3,'M',0),(3,'J',0),(3,'J',0),(3,'A',0),(3,'S',0),(3,'O',0),(3,'N',0),(3,'D',1);
--Query
WITH A
AS ( SELECT *,
CASE Year_Month
WHEN 'J' THEN 1
WHEN 'F' THEN 2
WHEN 'M' THEN 3
WHEN 'A' THEN 4
WHEN 'M' THEN 5
WHEN 'Ju' THEN 6
WHEN 'Jl' THEN 7
WHEN 'A' THEN 8
WHEN 'S' THEN 9
WHEN 'O' THEN 10
WHEN 'N' THEN 11
WHEN 'D' THEN 12
END
AS MonthNumber
FROM #Temp0 ),
StartingPoints
AS ( SELECT ID,
Year_Month,
MonthNumber,
Status
FROM A
WHERE NOT EXISTS
(
SELECT 1
FROM A
AS B
WHERE B.ID=A.ID
AND B.Status=A.Status-1
) ),
MonthRanking
AS ( SELECT A.*,
ROW_NUMBER( ) OVER( PARTITION BY A.ID ORDER BY A.MonthNumber )
AS rownum
FROM A
INNER JOIN
(
SELECT ID,
MAX( MonthNumber )+1
AS StartOfLastGroup
FROM StartingPoints
GROUP BY ID
)
AS B
ON A.ID=B.ID
AND A.MonthNumber>=B.StartOfLastGroup )
SELECT *
FROM MonthRanking
WHERE rownum=1;
Results:
If Month Names are recorded in Full as in July, June then this would work as well:
WITH StartingPoints
AS (SELECT ID,
Year_Month,
MonthNUmber = MONTH('01-'+Year_Month+'-2010'),
Status
FROM #Temp0
WHERE NOT EXISTS
(
SELECT 1
FROM #Temp0 AS B
WHERE B.ID = #Temp0.ID
AND B.Status = #Temp0.Status - 1
)),
MonthRanking
AS (SELECT A.*,
ROW_NUMBER() OVER(PARTITION BY A.ID ORDER BY MONTH('01-'+A.Year_Month+'-2010')) AS rownum
FROM #Temp0 AS A
INNER JOIN
(
SELECT ID,
MAX(MonthNumber) + 1 AS StartOfLastGroup
FROM StartingPoints
GROUP BY ID
) AS B ON A.ID = B.ID
AND MONTH('01-'+A.Year_Month+'-2010') >= B.StartOfLastGroup)
SELECT *
FROM MonthRanking
WHERE rownum = 1;
Results:
And if we assume that the data is as Iamdave assumes then it simply like so:
WITH StartingPoints
AS (SELECT ID,
Year_Month,
Status
FROM #Temp0
WHERE NOT EXISTS
(
SELECT 1
FROM #Temp0 AS B
WHERE B.ID = #Temp0.ID
AND B.Status = #Temp0.Status - 1
)),
MonthRanking
AS (SELECT A.*,
ROW_NUMBER() OVER(PARTITION BY A.ID ORDER BY Year_Month) AS rownum
FROM #Temp0 AS A
INNER JOIN
(
SELECT ID,
MAX(Year_Month) + 1 AS StartOfLastGroup
FROM StartingPoints
GROUP BY ID
) AS B ON A.ID = B.ID
AND A.Year_Month >= B.StartOfLastGroup)
SELECT *
FROM MonthRanking
WHERE rownum = 1;
Results:

You can do this with a couple derived tables that stack two window functions on top of one another (which can't be done in the same select). I have assumed that your data is slightly different to the table you have provided, based on the column names in your query. If they are not as I have them below, I strongly recommend having a look at how you store your data:
declare #t table(ID int, YearMonth int,StatusValue bit);
insert into #t values (1,201501,1),(1,201502,0),(1,201503,0),(1,201504,1),(1,201505,0),(1,201506,1),(1,201507,0),(1,201508,0),(1,201509,1),(1,201510,1),(1,201511,1),(1,201512,1),(2,201601,1),(2,201602,0),(2,201603,1),(2,201604,0),(2,201605,1),(2,201606,0),(2,201607,1),(2,201608,0),(2,201609,1),(2,201610,0),(2,201611,1),(2,201612,1),(3,201701,0),(3,201702,0),(3,201703,0),(3,201704,0),(3,201705,0),(3,201706,0),(3,201707,0),(3,201708,0),(3,201709,0),(3,201710,0),(3,201711,0),(3,201712,1);
with c as
(
select ID
,YearMonth
,StatusValue
,case when StatusValue = 1
and lead(StatusValue,1,1) over (partition by ID
order by YearMonth desc) = 0
then 1
else 0
end as c
from #t
), sc as
(
select ID
,YearMonth
,StatusValue
,sum(c) over (partition by ID order by YearMonth desc) as sc
from c
where c = 1
)
select ID
,YearMonth
,StatusValue
from sc
where sc = 1
order by ID;
Output:
+----+-----------+-------------+
| ID | YearMonth | StatusValue |
+----+-----------+-------------+
| 1 | 201509 | 1 |
| 2 | 201611 | 1 |
| 3 | 201712 | 1 |
+----+-----------+-------------+

how can I get all ids starting from a given id recursively in a postgresql table that references itself?

the title may not be very clear so let's consider this example (this is not my code, just taking this example to model my request)
I have a table that references itself (like a filesystem)
id | parent | name
----+----------+-------
1 | null | /
2 | 1 | home
3 | 2 | user
4 | 3 | bin
5 | 1 | usr
6 | 5 | local
Is it possible to make a sql request so if I choose :
1 I will get a table containing 2,3,4,5,6 (because this is the root) so matching :
/home
/home/user
/home/user/bin
/usr
etc...
2 I will get a table containing 3,4 so matching :
/home/user
/home/user/bin
and so on

Use recursive common table expression. Always starting from the root, use an array of ids to get paths for a given id in the WHERE clause.
For id = 1:
with recursive cte(id, parent, name, ids) as (
select id, parent, name, array[id]
from my_table
where parent is null
union all
select t.id, t.parent, concat(c.name, t.name, '/'), ids || t.id
from cte c
join my_table t on c.id = t.parent
)
select id, name
from cte
where 1 = any(ids) and id <> 1
id | name
----+-----------------------
2 | /home/
5 | /usr/
6 | /usr/local/
3 | /home/user/
4 | /home/user/bin/
(5 rows)
For id = 2:
with recursive cte(id, parent, name, ids) as (
select id, parent, name, array[id]
from my_table
where parent is null
union all
select t.id, t.parent, concat(c.name, t.name, '/'), ids || t.id
from cte c
join my_table t on c.id = t.parent
)
select id, name
from cte
where 2 = any(ids) and id <> 2
id | name
----+-----------------------
3 | /home/user/
4 | /home/user/bin/
(2 rows)
Bidirectional query
The question is really interesting. The above query works well but is inefficient as it parses all tree nodes even when we're asking for a leaf. The more powerful solution is a bidirectional recursive query. The inner query walks from a given node to top, while the outer one goes from the node to bottom.
with recursive outer_query(id, parent, name) as (
with recursive inner_query(qid, id, parent, name) as (
select id, id, parent, name
from my_table
where id = 2 -- parameter
union all
select qid, t.id, t.parent, concat(t.name, '/', q.name)
from inner_query q
join my_table t on q.parent = t.id
)
select qid, null::int, right(name, -1)
from inner_query
where parent is null
union all
select t.id, t.parent, concat(q.name, '/', t.name)
from outer_query q
join my_table t on q.id = t.parent
)
select id, name
from outer_query
where id <> 2; -- parameter

Creating clusters of related columns

I have a table named Stores with columns:
StoreCode NVARCHAR(10),
OldStoreCode NVARCHAR(10)
Here is a sample of my data:
| StoreCode | OldStoreCode |
|-----------|--------------|
| A | B |
| B | A |
| D | E |
| E | F |
| M | K |
| J | K |
| K | L |
|-----------|--------------|
I want to create clusters of related Stores. Related store means there is a one way relation between StoreCodes and OldStoreCodes.
Expected result table:
| StoreCode | ClusterId |
|-----------|-----------|
| A | 1 |
| B | 1 |
| D | 2 |
| E | 2 |
| F | 2 |
| M | 3 |
| K | 3 |
| J | 3 |
| L | 3 |
|-----------|-----------|
There is no maximum number hops. There may be a StoreCode A which has a OldStoreCode B, which has a OldStoreCode C, which has a OldStoreCode D etc.
How can I cluster stores like this?

Try it like this:
EDIT: With changes by OP taken from comment
DECLARE #tbl TABLE(ID INT IDENTITY, StoreCode VARCHAR(100),OldStoreCode VARCHAR(100));
INSERT INTO #tbl VALUES
('A','B'),('B','A'),('D','E'),('E','F'),('M','K'),('J','K'),('K','L');
WITH Related AS
(
SELECT DISTINCT t1.ID,Val
FROM #tbl AS t1
INNER JOIN #tbl AS t2 ON t1.StoreCode=t2.StoreCode
OR t1.OldStoreCode=t2.OldStoreCode
OR t1.OldStoreCode=t2.StoreCode
OR t1.StoreCode=t2.OldStoreCode
CROSS APPLY(SELECT DISTINCT Val
FROM
(VALUES(t1.StoreCode),(t2.StoreCode),(t1.OldStoreCode),(t2.OldStoreCode)) AS A(Val)
) AS valsInCols
)
,ClusterKeys AS
(
SELECT r1.ID
,(
SELECT r2.Val AS [*]
FROM Related AS r2
WHERE r2.ID=r1.ID
ORDER BY r2.Val
FOR XML PATH('')
) AS ClusterKey
FROM Related AS r1
GROUP BY r1.ID
)
,ClusterIds AS
(
SELECT ClusterKey
,MIN(ID) AS ID
FROM ClusterKeys
GROUP BY ClusterKey
)
SELECT r.ID
,r.Val
FROM ClusterIds c
INNER JOIN Related r ON c.ID = r.ID
The result
ID Val
1 A
1 B
3 D
3 E
3 F
5 J
5 K
5 L
5 M

This should do it:
SAMPLE DATA:
IF OBJECT_ID('tempdb..#Temp1') IS NOT NULL
BEGIN
DROP TABLE #Temp1;
END;
CREATE TABLE #Temp1(StoreCode NVARCHAR(10)
, OldStoreCode NVARCHAR(10));
INSERT INTO #Temp1(StoreCode
, OldStoreCode)
VALUES
('A'
, 'B'),
('B'
, 'A'),
('D'
, 'E'),
('E'
, 'F'),
('M'
, 'K'),
('J'
, 'K'),
('K'
, 'L');
QUERY:
;WITH A -- get all distinct new and old storecodes
AS (
SELECT StoreCode
FROM #Temp1
UNION
SELECT OldStoreCode
FROM #Temp1),
B -- give a unique number id to each store code
AS (SELECT rn = RANK() OVER(ORDER BY StoreCode)
, StoreCode
FROM A),
C -- combine the store codes and the unique number id's in one table
AS (SELECT b2.rn AS StoreCodeID
, t.StoreCode
, b1.rn AS OldStoreCodeId
, t.OldStoreCode
FROM #Temp1 AS t
LEFT OUTER JOIN B AS b1 ON t.OldStoreCode = b1.StoreCode
LEFT OUTER JOIN B AS b2 ON t.StoreCode = b2.StoreCode),
D -- assign a row number for each entry in the data set
AS (SELECT rn = RANK() OVER(ORDER BY StoreCode)
, *
FROM C),
E -- derive first and last store in the path
AS (SELECT FirstStore = d2.StoreCode
, LastStore = d1.OldStoreCode
, GroupID = d1.OldStoreCodeId
FROM D AS d1
RIGHT OUTER JOIN D AS d2 ON d1.StoreCodeID = d2.OldStoreCodeId
AND d1.rn - 1 = d2.rn
WHERE d1.OldStoreCode IS NOT NULL) ,
F -- get the stores wich led to the last store with one hop
AS (SELECT C.StoreCode
, E.GroupID
FROM E
INNER JOIN C ON E.LastStore = C.OldStoreCode)
-- combine to get the full grouping
SELECT A.StoreCode, ClusterID = DENSE_RANK() OVER (ORDER BY A.GroupID) FROM (
SELECT C.StoreCode,F.GroupID FROM C INNER JOIN F ON C.OldStoreCode = F.StoreCode
UNION
SELECT * FROM F
UNION
SELECT E.LastStore,E.GroupID FROM E) AS A ORDER BY StoreCode, ClusterID
RESULTS:

Selecting a row if it has all related options in a related table

Given a table definition:
Articles:
art_id | name
-------|--------------
1 | article1
2 | article2
3 | article3
Tags:
tag_id | description
-------|--------------
1 | Scientific
2 | Long
3 | Short
article_tags:
art_id | tag_id
-------|---------
1 | 1
1 | 2
2 | 1
2 | 3
3 | 1
3 | 2
3 | 3
The question is How to select all articles that are BOTH Scientific and Short?
Please note, it should be general for [2.N) tag combinations...

You can use the following query to get the result:
select a.art_id, a.name
from articles a
inner join article_tags at
on a.art_id = at.art_id
inner join tags t
on at.tag_id = t.tag_id
where t.description in ('Short', 'Scientific') -- tags here
group by a.art_id, a.name
having count(distinct t.tag_id) = 2 -- total count of tags here
See SQL Fiddle with Demo
Or this could be written:
select a.art_id, a.name
from articles a
inner join article_tags at
on a.art_id = at.art_id
inner join tags t
on at.tag_id = t.tag_id
group by a.art_id, a.name
having
sum(case when t.description = 'Short' then 1 else 0 end) >= 1 and
sum(case when t.description = 'Scientific' then 1 else 0 end) >= ;
See SQL Fiddle with Demo.
If you just want to return the article id, then you could just query the article_tag table:
select art_id
from article_tags
where tag_id in (1, 3)
group by art_id
having count(distinct tag_id) = 2
See SQL Fiddle with Demo

SELECT *
FROM articles
WHERE art_id IN
(
SELECT art_id
FROM article_tags
GROUP BY art_id
HAVING COUNT(art_id) > 1
)