Question

I have a table with values like the following

Name     DatePurchased     QuantityPurchased

A      2/3/2012             1
A      2/4/2012             1 
A      2/5/2012             2
B      2/2/2012             1
B      2/3/2012             2

I want to output the following

Name        DatePurchased         QuantityPurchased

A            2/3/2012               1
A            2/4/2012               2      // as I have purchased 2 upto this date
A            2/5/2012               4      // as I have purchased 4 upto this date
B            2/2/2012               1
B            2/3/2012               3

My query

SELECT Name,  `DatePurchased` , SUM(QuantityPurchased) 
FROM table1
GROUP BY DatePurchased

does not do the math right. I know whats wrong but can’t figure out the solution.

Thanks

Answer 1

Try:

SELECT t1.Name, t1.DatePurchased, SUM(t2.QuantityPurchased)
FROM table1 t1
LEFT JOIN table1 t2
ON t1.Name=t2.Name AND t1.DatePurchased >= t2.DatePurchased
GROUP BY Name, t1.DatePurchased

This joins table1 to itself within Name and such that t1s date is always at least t2s date, and sums up t2s QuantityPurchased (for each name,date in t1).

(Try performing the same query with SELECT *, and without the SUM and GROUP BY to see the joined table. Then the SUM and GROUP BY will become clear).