Reputation:
I am learning window functions in sql server. I am using AdventrueWorks2012 database for practice. I want to calculate total number of sales and purchases for each item in the store.
The classic solution can be like
SELECT ProductID,
Quantity,
(SELECT Count(*)
FROM AdventureWorks.Purchasing.PurchaseOrderDetail
WHERE PurchaseOrderDetail.ProductID = p.ProductID) TotalPurchases,
(SELECT Count(*)
FROM AdventureWorks.Sales.SalesOrderDetail
WHERE SalesOrderDetail.ProductID = p.ProductID) TotalSales
FROM (SELECT DISTINCT ProductID,
Quantity
FROM AdventureWorks.Production.ProductInventory) p
Trying to convert to window functions gives me wrong results:
SELECT DISTINCT d.ProductID,
Quantity,
Count(d.ProductID)
OVER(
PARTITION BY d.ProductID) TotalPurchases,
Count(d2.ProductID)
OVER(
PARTITION BY d2.ProductID) TotalSales
FROM (SELECT DISTINCT ProductID,
Quantity
FROM AdventureWorks.Production.ProductInventory) p
INNER JOIN AdventureWorks.Purchasing.PurchaseOrderDetail d
ON p.ProductID = d.ProductID
INNER JOIN AdventureWorks.Sales.SalesOrderDetail d2
ON p.ProductID = d2.ProductID
ORDER BY d.ProductID
Why this is wrong? How can I correct it?
Upvotes: 1
Views: 1346
Reputation: 1464
You should change INNER JOIN
to LEFT JOIN
Because when you inner join, result will miss productid which from ProductInventory
table does not have PurchaseOrderDetail
or SalesOrderDetail
.
Upvotes: 1