August 29, 2016 Leave a comment
This article was recently published on dev.getroadmap.com :
If there is one thing I pretty much hate doing in T-SQL it’s PIVOT and UNPIVOT . Even though I used it a few times in the last couple of years, it’s an adventure to find out how it works every time. And I know a lot of people struggle with this part of T-SQL, so let’s take a look at a (hopefully) simple example.
Key-Value pairs
The idea of using key-value pairs to store data isn’t new, but with the rapid development of cloud solutions like Azure and the hype around NoSQL databases, using key-value pairs to store data got a big boost. Especially developers (in my experience) love using key-value pair to store their data, because it’s easy for them to consume the data in an application. But it gives the database professional an extra challenge because we’re used to retrieve columns with values instead of a record per value. So how can we turn those key-value pairs into rows?
Flights
To show you how to change key-value pairs to columns, let’s create an example based on flights. If one of our customers needs to travel and they take the plane, there is some basic information we need to show them the flight status in the app. So let’s create a table to store that dataset:
CREATE TABLE dbo.Flights(
FlightId UNIQUEIDENTIFIER,
[Key] VARCHAR(255),
[Value] VARCHAR(255)
)
and insert a flight:
INSERT INTO dbo.Flights(FlightId, [Key], [Value])
VALUES
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'FlightId', '3074e015-62b2-4f76-a8b1-463c53cd79c5'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'AirlineIATACode', 'VY'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'FlightNumber', '8336'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'DepartureAirportCode', 'RTM'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'DepartureAirportName', 'Rotterdam The Hague Airport'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'DepartureDateTime', '2016-08-12 12:15'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'DepartureTerminal', '1'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'Gate', NULL),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'ArrivalAirportCode', 'BCN'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'ArrivalAirportName', 'Barcelona El Prat'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'ScheduledArrivalDateTime', '2016-08-12 14:14'),
('2A0C8B05-682A-41CE-8516-C6070CD92851', 'FlightIsCancelled', '0')
So now we have a dataset that consists of 12 key-value pairs:
Converting to columns
But when you’re running a query you would like to have columns instead of rows. So by using a PIVOT statement, you can convert these rows into columns:
SELECTFlightId,
AirlineIATACode,
FlightNumber,
DepartureAirportCode,
DepartureAirportName,
DepartureDateTime,
DepartureTerminal,
Gate,
ArrivalAirportCode,
ArrivalAirportName,
ScheduledArrivalDateTime,
FlightIsCancelled
FROM
(
SELECT
FlightId,
[Key],
[Value]
FROM dbo.Flights
) AS SourceTable
PIVOT
(
MIN([Value]) --Needs to be an aggregate function
FOR [Key] IN
(
AirlineIATACode,
FlightNumber,
DepartureAirportCode,
DepartureAirportName,
DepartureDateTime,
DepartureTerminal,
Gate,
ArrivalAirportCode,
ArrivalAirportName,
ScheduledArrivalDateTime,
FlightIsCancelled
)
) AS PivotOutput
Because the keys are always the same for flights, I can add those Key-names in the PIVOT statement. But because the PIVOT statement needs an aggregate function to retrieve the value, I needed to add the MIN() function but this doesn’t change the output in any way.
Now we have the same dataset, but converted into rows:
So from now on, converting key-value pairs to columns shouldn’t be a problem anymore!