Sony Francis @EY (10/1/2012)

---

May be this query will help your problem


CREATE TABLE #Temp (Name varchar(100), value int)

INSERT INTO #Temp
SELECT 'Day1', 12
UNION
SELECT 'Day2' ,6
UNION
SELECT 'Day3' , 10
UNION
SELECT 'Day4' ,8
UNION
SELECT 'Day5' ,4

;with cte as
(
select *, ROW_NUMBER() over(PARTITION by null order by Name ) as Id
FROM #Temp
)


SELECT b.Name , SUM(A.value ), B.Id
FROM cte A
INNER JOIN cte B on A.Id <= B.Id
GROUP BY B.Name , b.id

I believe Jeff talks about this "triangular join" in the article I referenced.

diamondgm (10/1/2012)

---

I don't think that its a triangular join - no subquery really.
It looks like a moficifation of the CROSS JOIN method of running total computation.
Instead of a filtered CROSS JOIN, the INNER JOIN condition is expanded.
This is something that Jeff did not cover in his article as far as I can remember.

This is my understanding; I'm open to correction.