Text Size: Normal / Large

Home → Documentation → Manuals → PostgreSQL 7.4

PostgreSQL 7.4.16 Documentation
Prev	Fast Backward	Chapter 9. Functions and Operators	Fast Forward	Next

9.15. Aggregate Functions

Aggregate functions compute a single result value from a set of input values. Table 9-43 shows the built-in aggregate functions. The special syntax considerations for aggregate functions are explained in Section 4.2.7. Consult Section 2.7 for additional introductory information.

Table 9-43. Aggregate Functions

Function	Argument Type	Return Type	Description
`avg(expression)`	`smallint`, `integer`, `bigint`, `real`, `double precision`, `numeric`, or `interval`	`numeric` for any integer type argument, `double precision` for a floating-point argument, otherwise the same as the argument data type	the average (arithmetic mean) of all input values
`count(*)`		`bigint`	number of input values
`count(expression)`	any	`bigint`	number of input values for which the value of `expression` is not null
`max(expression)`	any numeric, string, or date/time type	same as argument type	maximum value of `expression` across all input values
`min(expression)`	any numeric, string, or date/time type	same as argument type	minimum value of `expression` across all input values
`stddev(expression)`	`smallint`, `integer`, `bigint`, `real`, `double precision`, or `numeric`	`double precision` for floating-point arguments, otherwise `numeric`	sample standard deviation of the input values
`sum(expression)`	`smallint`, `integer`, `bigint`, `real`, `double precision`, `numeric`, or `interval`	`bigint` for `smallint` or `integer` arguments, `numeric` for `bigint` arguments, `double precision` for floating-point arguments, otherwise the same as the argument data type	sum of `expression` across all input values
`variance`(`expression`)	`smallint`, `integer`, `bigint`, `real`, `double precision`, or `numeric`	`double precision` for floating-point arguments, otherwise `numeric`	sample variance of the input values (square of the sample standard deviation)

It should be noted that except for count, these functions return a null value when no rows are selected. In particular, sum of no rows returns null, not zero as one might expect. The function coalesce may be used to substitute zero for null when necessary.

Note: Users accustomed to working with other SQL database management systems may be surprised by the performance characteristics of certain aggregate functions in PostgreSQL when the aggregate is applied to the entire table (in other words, no WHERE clause is specified). In particular, a query like
SELECT min(col) FROM sometable;
will be executed by PostgreSQL using a sequential scan of the entire table. Other database systems may optimize queries of this form to use an index on the column, if one is available. Similarly, the aggregate functions max() and count() always require a sequential scan if applied to the entire table in PostgreSQL.
PostgreSQL cannot easily implement this optimization because it also allows for user-defined aggregate queries. Since min(), max(), and count() are defined using a generic API for aggregate functions, there is no provision for special-casing the execution of these functions under certain circumstances.
Fortunately, there is a simple workaround for min() and max(). The query shown below is equivalent to the query above, except that it can take advantage of a B-tree index if there is one present on the column in question.
SELECT col FROM sometable ORDER BY col ASC LIMIT 1;
A similar query (obtained by substituting DESC for ASC in the query above) can be used in the place of max()).
Unfortunately, there is no similarly trivial query that can be used to improve the performance of count() when applied to the entire table.

Prev	Home	Next
Array Functions and Operators	Up	Subquery Expressions

User Comments

Steinar H. Gunderson <sgunderson AT bigfoot.com>

27 Nov 2004 14:15:47

This is a bit of a hack, but I was surprised COUNT(expression) counted non-NULL instead of true values (as in COUNT(a=b) for counting all rows where a=b). The simple solution is doing COUNT(a=b OR NULL), which will be non-NULL if a=b (TRUE OR NULL is true), and NULL otherwise (FALSE OR NULL is NULL). :-)

Shachar Shemesh <pgdoc AT shemesh.biz>

16 Jan 2005 14:25:29

For info on writing your own aggregate functions, see 33.9

Add Comment

Please use this form to add your own comments regarding your experience with particular features of PostgreSQL, clarifications of the documentation, or hints for other users. Please note, this is not a support forum, and your IP address will be logged. If you have a question or need help, please see the faq, try a mailing list, or join us on IRC. Note that submissions containing URLs or other keywords commonly found in 'spam' comments may be silently discarded. Please contact the webmaster if you think this is happening to you in error.

In order to submit a comment, you must have a community account.

* denotes required field