Is it possible to have an 'ACTUAL' row followed by an 'ESTIMATED' row?

If not, it might be easier to just query for the most recent row (using Cory's excellent setup...

Are you using MySQL or SQL Server?

If the former, you might be better off using a MySql forum since this is a SQL Server forum. There are "a few differences"...

I second Michael.

And additionally, why don't you use STRG+H in SSMS and replace every single quotation mark with a double in the required section?

As a side note:

dbo.fnReadField(description, ' + ''''...

Let's try to clarify the business requirement:

What would be the expected output if you'd add a row with id=10 and value =100? What rows will be ignored? (my assumption: 3,4,7,8,...

Glad I could help 😀

You could use your current query as a subquery/CTE and use ROW_NUMBER() OVER(PARTITION BY rank_result_column ORDER BY order_criteria).

If this "hint" doesn't provide the help you need please post table DDL,...

I would use the "quirky update" method to set a flag in a separate column (delete flag).

For details on how to use this method please see Jeffs great article[/url].

I consider...

Is the following code close to what you're looking for?

I'm using a self referencing join based on ROW_NUMBER.

;WITH cte AS

(

SELECT *,

ROW_NUMBER() OVER(PARTITION BY opportunityid,[MONTH] ORDER BY pricebookHistoryID ) AS...

Would something like the following get the result you require?

I used a CTE to get the rows with OrderDate within the last three month as well as the total count...

Please clarify your requirement: Do you need to to show just the total number of counts within the last three months (then it wouldn't make sense to display each month...

Would you please clarify the content of dbo.pricehistory as well as your expected result?

Which one do you consider as a "numeric digit"?

I see two valid numeric values in each row (e.g. 90001415 and 4.000 are both valid numeric values...). Is there a deterministic...

You know, I consider this forum...

I'm not sure what the DDL for the transaction table looks like... Table def, sample data and expected result would help a lot here...

I'd probably use "ROW_NUMBER() OVER(PARTITION BY CustomerID...

I'm sure one of the...