WITH yourCTEname

AS

(SELECT

CustID

,MAX(OrderDate) AS [Last Order]

FROM Orders

WHERE CustID

GROUP BY CustID)

SELECT

ID,

CustName,

FROM Customers

LEFT JOIN...

....but something tells me there's an easier way.

Include another set of joins...

SELECT

pr.parentprodnodeid

,pr.childprodnodeid

,po.prodnodeid

,pc.sysclassname AS [Parent Class]

,cc.sysclassname AS [Child Class]

FROM

productrelationships pr

LEFT JOIN productobjects po

ON pr.parentprodnodeid = po.prodnodeid

LEFT JOIN productclasses pc

ON po.prodclassid = pc.prodclassid

LEFT JOIN productobjects co

ON pr.childprodnodeid...

The "Hitchhikers Guide to SQL Reporting Services 2000" is a great book. Although it's for 2000, a lot of it is still applicable.

http://www.sqlreportingservices.net/

Yes, a table variable would in deed be faster.

[p]WITH

problem (problem_name, kount)

AS (SELECT

problem_name = pr.name

...

Well, poo! I wasn't thinking of it that way. Of course, if you need info from both CTE's then you'll need to join them in the last select.

No, CTE's are only available to the statement immediately following them.

you will want to use a table variable instead.

But I like random, it's rather difficult to disprove random results... :hehe:

Change "transactionDate" to "Spend DESC" in my code and you should get what you need.:D

First you want to get away from procedural thinking. SQL works best set based. Also, post the DDL for your tables, it makes it easier for people to respond with...

Andras is correct. And easy way to do this would be to throw your PIVOT into a CTE and then you can call it twice in the union.

You simply put the name of your dataset or group in place of SCOPE. This will return the number of rows in the dataset or group. You do not need...

It's a little confusing. I'm still not 100% on them (pivot and unpivot) but it's beginning to get clear. DONT ASK ME TO EXPLAIN.. *laughs*

But as you can see the...

Simply put, each row from one table is returned with each row from the second or a Cartesian Product is the result set. So if each table has 10 rows,...