SQL Recursive CTE unexpectedly returns alternating sets - sql

I am trying to get the use recursive CTE to repeat the same pattern over and over, resetting when "Scenario" increases in value. RowNumber repeats 1-21 (as desired), but whenever "Scenario" is an even number, there are too few items in the "Vals" column to feed into "Value". I can't figure out which part of the code is causing me to be 1 short for only even Scenarios.
Below are the results of the code I'm using at the bottom.
Scenario RowNumber Value Vals
1 1 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C
1 2 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C
1 3 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C
1 4 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C
1 5 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C
1 6 A A,A,A,A,A,A,A,A,A,A,A,A,A,B,C
1 7 A A,A,A,A,A,A,A,A,A,A,A,A,B,C
1 8 A A,A,A,A,A,A,A,A,A,A,A,B,C
1 9 A A,A,A,A,A,A,A,A,A,A,B,C
1 10 A A,A,A,A,A,A,A,A,A,B,C
1 11 A A,A,A,A,A,A,A,A,B,C
1 12 A A,A,A,A,A,A,A,B,C
1 13 A A,A,A,A,A,A,B,C
1 14 A A,A,A,A,A,B,C
1 15 A A,A,A,A,B,C
1 16 A A,A,A,B,C
1 17 A A,A,B,C
1 18 A A,B,C
1 19 A B,C
1 20 B C
1 21 C
2 1 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,C
2 2 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,C
2 3 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,C
2 4 A A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,C
2 5 A A,A,A,A,A,A,A,A,A,A,A,A,B,B,C
2 6 A A,A,A,A,A,A,A,A,A,A,A,B,B,C
2 7 A A,A,A,A,A,A,A,A,A,A,B,B,C
2 8 A A,A,A,A,A,A,A,A,A,B,B,C
2 9 A A,A,A,A,A,A,A,A,B,B,C
2 10 A A,A,A,A,A,A,A,B,B,C
2 11 A A,A,A,A,A,A,B,B,C
2 12 A A,A,A,A,A,B,B,C
2 13 A A,A,A,A,B,B,C
2 14 A A,A,A,B,B,C
2 15 A A,A,B,B,C
2 16 A A,B,B,C
2 17 A B,B,C
2 18 B B,C
2 19 B C
2 20 C
2 21 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,C
3 1 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C,C
3 2 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C,C
3 3 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C,C
3 4 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,C,C
3 5 A A,A,A,A,A,A,A,A,A,A,A,A,A,B,C,C
3 6 A A,A,A,A,A,A,A,A,A,A,A,A,B,C,C
3 7 A A,A,A,A,A,A,A,A,A,A,A,B,C,C
3 8 A A,A,A,A,A,A,A,A,A,A,B,C,C
3 9 A A,A,A,A,A,A,A,A,A,B,C,C
3 10 A A,A,A,A,A,A,A,A,B,C,C
3 11 A A,A,A,A,A,A,A,B,C,C
3 12 A A,A,A,A,A,A,B,C,C
3 13 A A,A,A,A,A,B,C,C
3 14 A A,A,A,A,B,C,C
3 15 A A,A,A,B,C,C
3 16 A A,A,B,C,C
3 17 A A,B,C,C
3 18 A B,C,C
3 19 B C,C
3 20 C C
3 21 C
4 1 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,B,C
4 2 A A,A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,B,C
4 3 A A,A,A,A,A,A,A,A,A,A,A,A,A,B,B,B,C
4 4 A A,A,A,A,A,A,A,A,A,A,A,A,B,B,B,C
4 5 A A,A,A,A,A,A,A,A,A,A,A,B,B,B,C
4 6 A A,A,A,A,A,A,A,A,A,A,B,B,B,C
4 7 A A,A,A,A,A,A,A,A,A,B,B,B,C
4 8 A A,A,A,A,A,A,A,A,B,B,B,C
4 9 A A,A,A,A,A,A,A,B,B,B,C
4 10 A A,A,A,A,A,A,B,B,B,C
4 11 A A,A,A,A,A,B,B,B,C
4 12 A A,A,A,A,B,B,B,C
4 13 A A,A,A,B,B,B,C
4 14 A A,A,B,B,B,C
4 15 A A,B,B,B,C
4 16 A B,B,B,C
4 17 B B,B,C
4 18 B B,C
4 19 B C
4 20 C
This is the code I used to generate the above sample. Where am I going wrong?
CREATE TABLE #temp3
(
Scenario INT
,Vals VARCHAR(64)
,LEN INT
)
;
WITH vals AS
(
SELECT
v.*
FROM
(VALUES ('A'), ('B'), ('C')) v(x)
),
CTE AS
(
SELECT CAST('A' AS VARCHAR(MAX)) AS STR, 0 AS LEN
UNION ALL
SELECT (CTE.STR + ',' + vals.x), CTE.LEN + 1
FROM
CTE
JOIN vals
ON vals.x >= RIGHT(CTE.STR, 1)
WHERE CTE.LEN < 19
)
INSERT INTO #temp3
SELECT
ROW_NUMBER() OVER(ORDER BY STR + ',C') AS Scenario
,STR + ',C' AS Vals
,LEN
FROM
CTE
WHERE
STR + 'C' LIKE '%B%'
AND LEN = 19
;
-- Split strings created above into individual characters
WITH cte(Scenario, Value, Vals) AS
(
SELECT
Scenario
,CAST(LEFT(Vals, CHARINDEX(',',Vals+',')-1) AS VARCHAR(10)) AS Value
,STUFF(Vals, 1, CHARINDEX(',',Vals+','), '') AS Vals
FROM #temp3
UNION ALL
SELECT
Scenario
,CAST(LEFT(Vals, CHARINDEX(',',Vals+',')-1) AS VARCHAR(10))
,STUFF(Vals, 1, CHARINDEX(',',Vals+','), '')
FROM cte
WHERE Vals > ''
)
SELECT
Scenario
,ROW_NUMBER() OVER (PARTITION BY Scenario ORDER BY Scenario) RowNumber
,Value
,Vals
FROM cte t

I'm not exactly sure what the problem you are describing is, but the ROW_NUMBER() should use an ORDER BY clause that completely orders the rows in each partition.
When you use "PARTITION BY Scenario ORDER BY Scenario" the order in which the ROW_NUMBER() values are assigned is undefined. Try something like
WITH cte(Scenario, depth, Value, Vals) AS
(
SELECT
Scenario, 0 depth
,CAST(LEFT(Vals, CHARINDEX(',',Vals+',')-1) AS VARCHAR(10)) AS Value
,STUFF(Vals, 1, CHARINDEX(',',Vals+','), '') AS Vals
FROM #temp3
UNION ALL
SELECT
Scenario, depth+1
,CAST(LEFT(Vals, CHARINDEX(',',Vals+',')-1) AS VARCHAR(10))
,STUFF(Vals, 1, CHARINDEX(',',Vals+','), '')
FROM cte
WHERE Vals > ''
)
SELECT
Scenario
,depth
,ROW_NUMBER() OVER (PARTITION BY Scenario ORDER BY depth ) RowNumber
,Value
,Vals
FROM cte t

Related

row_number() but only increment value after a specific value in a column

Query: SELECT (row_number() OVER ()) as grp, * from tbl
Edit: the rows below are returned by a pgrouting shortest path function and it does have a sequence.
seq grp id
1 1 8
2 2 3
3 3 2
4 4 null
5 5 324
6 6 82
7 7 89
8 8 null
9 9 1
10 10 2
11 11 90
12 12 null
How do I make it so that the grp column is only incremented after a null value on id - and also keep the same order of rows
seq grp id
1 1 8
2 1 3
3 1 2
4 1 null
5 2 324
6 2 82
7 2 89
8 2 null
9 3 1
10 3 2
11 3 90
12 3 null
demo:db<>fiddle
Using a cumulative SUM aggregation is a possible approach:
SELECT
SUM( -- 2
CASE WHEN id IS NULL THEN 1 ELSE 0 END -- 1
) OVER (ORDER BY seq) as grp,
id
FROM mytable
If the current (ordered!) value is NULL, then make it 1, else 0. Now you got a bunch of zeros, delimited by a 1 at each NULL record. If you'd summerize these values cumulatively, at each NULL record, the sum increased.
Execution of the cumulative SUM() using window functions
This yields:
0 8
0 3
0 2
1 null
1 324
1 82
1 89
2 null
2 1
2 2
2 90
3 null
As you can see, the groups start with the NULL records, but you are expecting to end it.
This can be achieved by adding another window function: LAG(), which moves the records to the next row:
SELECT
SUM(
CASE WHEN next_id IS NULL THEN 1 ELSE 0 END
) OVER (ORDER BY seq) as grp,
id
FROM (
SELECT
LAG(id) OVER (ORDER BY seq) as next_id,
seq,
id
FROM mytable
) s
The result is your expected one:
1 8
1 3
1 2
1 null
2 324
2 82
2 89
2 null
3 1
3 2
3 90
3 null

SQL: subset data: select id when time_id for id satisfy a condition from another column

I have a data (dt) in SQL like the following:
ID time_id act rd
11 1 1 1
11 2 4 1
11 3 7 0
12 1 8 1
12 2 2 0
12 3 4 1
12 4 3 1
12 5 4 1
13 1 4 1
13 2 1 0
15 1 3 1
16 1 8 0
16 2 8 0
16 3 8 0
16 4 8 0
16 5 8 0
and I want to take the subset of this data such that only ids (and their corresponding time_id, act, rd) that has time_id == 5 is retained. The desired output is the following
ID time_id act rd
12 1 8 1
12 2 2 0
12 3 4 1
12 4 3 1
12 5 4 1
16 1 8 0
16 2 8 0
16 3 8 0
16 4 8 0
16 5 8 0
I know I should use having clause somehow but have not been successful so far (returns me empty outputs). below is my attempt:
SELECT * FROM dt
GROUP BY ID
Having min(time_id) == 5;
This query:
select id from tablename where time_id = 5
returns all the ids that you want in the results.
Use it with the operator IN:
select *
from tablename
where id in (select id from tablename where time_id = 5)
You can use a correlated subquery with exists:
select t.*
from t
where exists (select 1 from t t2 where t2.id = t.id and t2.time_id = 5);
WITH temp AS
(
SELECT id FROM tab WHERE time_id = 5
)
SELECT * FROM tab t join temp tp on(t.id=tp.id);
check this query
select * from table t1 join (select distinct ID from table t where time_id = 5) t2 on t1.id =t2.id;

How to group/List all nodes of a undirected graph using teradata sql

I have data for many diff. set of undirected graphs in a table (like adjacent list relationship, one node is connected which all node) and I need to group all individual undirected graphs.
Eg: all nodes of the particular undirected graphs will be in a group & group name will be the min. of the node.
sel d.adj_node, min(d.adj_node) Over (Partition By a.node) as grp
table a
left join table b
on a.adj_node=b.node
left join table c
on b.adj_node=c.node
​left join table d
​on c.adj_node=d.node​;
Now, I am doing a self-join for 4,5 times and then on top that query doing partitioning it to get the desired output. But doing self-join 4 5 times is creating performance issue.
So, need some recursive sql, stored procedure or some other logic to do the same for all levels. Input Data & Required Output will be like this link Looking for some suggestions.
Input Table
node adj_node
1 2
2 1
2 3
2 5
2 6
2 7
3 2
3 4
4 3
4 5
4 6
4 7
5 2
5 4
6 2
6 4
6 8
7 2
7 4
8 6
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
10 11
11 10
11 13
11 14
12 13
12 14
13 11
13 12
13 14
14 11
14 12
14 13
10 10
11 11
12 12
13 13
14 14
Output
node grp
1 1
2 1
3 1
4 1
5 1
6 1
7 1
8 1
10 10
11 10
12 10
13 10
14 10
I just remembered that I did something similar before using updates on a temp table.
The best way to implement this would be a Stored Procedure with a loop in it:
CREATE VOLATILE TABLE vt_tab AS
(
SELECT DISTINCT NODE , adj_node, NODE AS grp
FROM tab AS t1
WHERE adj_node <> NODE
) WITH DATA
ON COMMIT PRESERVE ROWS
;
-- REPEAT this update UNTIL activity_count = 0
UPDATE vt_tab FROM
(
SELECT t2.NODE, MIN(t1.grp) AS mingrp
FROM vt_tab AS t1 JOIN vt_tab AS t2
ON t1.adj_node = t2.NODE
AND t1.grp < t2.grp
GROUP BY t2.NODE
) x
SET grp = mingrp
WHERE vt_tab.NODE = x.NODE
;
--get the final result
SEL DISTINCT NODE,grp
FROM vt_tab
ORDER BY 1
;
Recursion might be possible, but there's a high probability that it will blow your spool because you need repeated m:n joins and only the final Select allows to reduce the result rows.
A solution with a recursive CTE:
with cte as
(
select node as node, node as grp
from Tabl_1
Union all
select C.node, T.adj_node
from CTE C inner join Tabl_1 T on C.grp = T.node
where T.adj_node < C.grp
)
select node, MIN(grp) as grp
from cte
group by node
order by node
== EDIT 1 ==
Here is a new version to reflect your point.
with cte as
(
select node as node, node as grp, ',' + CAST(node as varchar(max)) + '-' + CAST(node as varchar(max)) + ',' as pair
from Tabl_1
Union all
select C.node, T.adj_node, C.pair + CAST(C.node as varchar(max)) + '-' + CAST(T.adj_node as varchar(max)) + ','
from CTE C inner join Tabl_1 T on C.grp = T.node
where C.pair not like '%,' + CAST(C.node as varchar(max)) + '-' + CAST(T.adj_node as varchar(max)) + ',%'
)
select node, MIN(grp) as grp
from cte
group by node
order by node

Select Data based on Sum of another columns value

I have a Table with Data as
RowIndex Id TicketCount
1 23 1
2 25 2
3 3 1
4 14 1
5 16 1
6 18 1
7 1 1
8 6 1
9 15 1 ===> at this row the sum of Ticket Count is 10
10 22 1
11 27 1
12 24 1
13 26 2
14 9 1
15 19 1
From this Data I want to Select All Records where The Sum of Ticket Count will be equal to 10(user input value)
In the Given data I want to Select all Records till Row Index 9.
Output should be:
RowIndex Id TicketCount
1 23 1
2 25 2
3 3 1
4 14 1
5 16 1
6 18 1
7 1 1
8 6 1
9 15 1
SQL Server 2008 doesn't have the cumulative sum function. I implement it using a correlated subquery:
select RowIndex, Id, TicketCount
from (select t.*,
(select sum(TicketCount)
from t t2
where t2.RowIndex <= t.RowIndex
) as cumTicketCount
from t
) t
where cumTicketCount <= 10;
In SQL Server 2012, you can phrase this using a window function:
select RowIndex, Id, TicketCount
from (select t.*, sum(TicketCount) over (order by RowIndex) as CumTicketCount
from t
) t
where cumTicketCount <= 10;
You can do it using recursive CTE:
WITH RCTE AS
(
SELECT *, TicketCount AS Total
FROM Table1
WHERE RowIndex = 1
UNION ALL
SELECT t.*, r.Total + t.TicketCount
FROM RCTE r
INNER JOIN Table1 t ON r.RowIndex + 1 = t.RowIndex
WHERE r.Total + t.TicketCount <= 10 --your input value
)
SELECT * FROM RCTE
SQLFiddle DEMO

Convert row to column using sql server 2008?

Table name is Looupvalue
id Ptypefield Value
1 1 D
2 1 E
3 1 F
4 1 G
5 1 H
6 2 FL
7 2 IF
8 2 VVS1
9 2 VVS2
10 2 VS1
11 2 VS2
12 3 0.50
13 3 1.00
14 3 1.50
15 3 2.00
16 4 Marquise
17 4 Round
18 4 Pear
19 4 Radiant
20 4 Princess
Lookupvalue table value convert roow to column depends on ptypefield
Like
id 1 id 2 id 3 id 4
1 D 6 fl 12 0.50 16 Marquise
2 E 7 If 13 1 17 Round....
3 F 8 vvs2 14 1.5
4 G 9 vvs2 15 2
5 H 10 vs1
11 vs2
Thanks
In your sample output, it is not clear why values from columns 1 and 2 would be related to columns 3 and 4. However, here is a possible solution:
;With RowNumbers As
(
Select Id, PTypeField, Value
, Row_Number() Over( Partition By PTypeField Order By Id ) As Rownum
From #Test
)
Select RowNum
, Min( Case When PTypeField = 1 Then Id End ) As Id
, Min( Case When PTypeField = 1 Then Value End ) As [1]
, Min( Case When PTypeField = 2 Then Id End ) As Id
, Min( Case When PTypeField = 2 Then Value End ) As [2]
, Min( Case When PTypeField = 3 Then Id End ) As Id
, Min( Case When PTypeField = 3 Then Value End ) As [3]
, Min( Case When PTypeField = 4 Then Id End ) As Id
, Min( Case When PTypeField = 4 Then Value End ) As [4]
From RowNumbers
Group By RowNum
If you wanted to dynamically generate the columns, the only way to do that in SQL is to use some fugly dynamic SQL. T-SQL was not designed for this sort of output and instead you should use a reporting tool or do the crosstabbing in a middle tier component or class.
This data schema looks like an EAV which would explain why retrieving the data you want is so difficult.