Sorry I’ve not posted anything in a couple weeks, I’ll back to it soon. I recently picked up Monster Hunter World on PC (already owned it on PS4) and been distracted by it (using it as an excuse to procrastinate) to get to a similar point in the game. I’ll get back to the SQL Server world (and finishing the T-SQL language extension for Prism) soon. I’ll aim to get part 2 on Pivots done by end of this week.
This is something that I’ve found a few people have missed and were surprised when I told them. This appears to have been snuck in with the Deprecated Database Engine Features in SQL Server 2016 under Features deprecated in a future version of SQL Server (note that in 2017 it is still under that heading and 2019 introduced no new deprecations). Specifically, from the documentation it states: The following SQL Server Database Engine features are supported in the next version of SQL Server, but will be deprecated in a later version. The specific version of SQL Server has not been…
Even before SQL Server 2012, ISNUMERIC was a function that you best avoided. Put quite simply, it’s just bad. It often provides results that are wrong, and it isn’t data type specific, something that’s actually really important when you have several numerical data types to use: int/bigint and the smaller ones, numeric/decimal, float/real and money. All of these data types behave very differently when converting from a varchar. Let’s start by looking at the results for ISNUMERIC: CREATE TABLE dbo.BadNumbers(Number varchar(20)); INSERT INTO dbo.BadNumbers VALUES(‘1’), (‘7.8’), (”), (‘1,020’), (‘1,079,190.7’), (‘17,1,68,11.0’), (‘1/3’), (‘1.236E7’), (‘2d6’); SELECT Number, ISNUMERIC(Number) AS IsNumber FROM dbo.BadNumbers;…
The title pretty much says it all here, when you’re writing T-SQL, the correct place to place the semicolon (;) is at the end of all you’re statements. Something I see quite often is batches where people don’t terminate any of their statements, but do start statements that use a CTE or Merge statement with a semicolon. This means you see syntax like the below: DECLARE @MyVariable int SET @MyVariable = 5 ;WITH MyCTE AS( SELECT *, ROW_NUMBER() OVER (PARTITION BY CONVERT(date,MyDate) ORDER BY MyDate DESC) AS RN FROM MyTable WHERE MyColumn = @MyVariable) SELECT * FROM MyCTE WHERE RN…
There are a couple of ways to Pivot your data in SQL Server. The first is by using PIVOT, and the other is using what is known as a cross-tab. Personally I prefer the latter as it’s more flexible and feel it’s easier to understand. I’ll explain in a later article why a cross-tab is more flexible, so I’ll be covering both methods in this article. In a addition to pivoting, you also have unpivoting. Again, you have two options here, using UNPIVOT or using a VALUES constructor. Again, I prefer the latter for the same reasons but I’ll be…