~drizzle-trunk/drizzle/development

Group By

========

 

The GROUP BY clause is used to extract only those records that fulfill a specified criterion.

 

SQL GROUP BY Syntax ::

 

        SELECT column_name, aggregate_function(column_name)

        FROM table_name

        WHERE column_name operator value

        GROUP BY column_name

 

        

**GROUP BY Clause Example**

 

The "Activities" table:

 

+---------+--------------+--------------+-------------+----------+

|Id       |ActivityDate  |ActivityType  |ActivityCost | userID   |

+=========+==============+==============+=============+==========+

| 1       |2011-01-02    | Sport        |45           |131       |

+---------+--------------+--------------+-------------+----------+

| 2       |2011-01-02    | Art          |10           |256       |

+---------+--------------+--------------+-------------+----------+

| 3       |2011-01-02    | Music        |25           |022       |

+---------+--------------+--------------+-------------+----------+

| 4       |2011-01-02    | Food         |125          |022       |

+---------+--------------+--------------+-------------+----------+

| 5       |2011-01-03    | Music        |40           |131       |

+---------+--------------+--------------+-------------+----------+

| 6       |2011-01-03    | Food         |20           |175       |

+---------+--------------+--------------+-------------+----------+

 

Running the following simple query::

 

        SELECT userID

        FROM activities

        GROUP BY userID;

 

Returns:

 

+---------+

| userID  |

+=========+

| 131     |

+---------+

| 256     |

+---------+

| 022     |

+---------+

| 175     |

+---------+

 

(This shows that GROUP BY accepts a column_name and consolidates like customer values.)

 

However, GROUP BY is much more powerful when used with an aggregate function. Let's say you want to find the total amount spent by each unique User.

 

You could use the following SQL statement: ::

        

        SELECT userID,SUM(ActivityCost) AS "Activity Total"

        FROM Activities

        GROUP BY userID;

 

The result-set will look like this:

 

======    ==============

userID    Activity Total

======    ==============

131       85             

256       10             

022       150            

175       20             

======    ==============

 

With the aggregate SUM() function, SQL can calculate how much each unique user has spent on activities over time.

 

We can also use the GROUP BY statement on more than one column, like this: ::

        

        SELECT userID,ActivityDate,SUM(ActivityCost) 

        FROM Activities

        GROUP BY userID,ActivityDate;