deﬁnes a new aggregate function. Some basic and commonly-used aggregate
functions are included with the distribution; they are documented in Section 9.20. If one deﬁnes new
types or needs an aggregate function not already provided, then
can be used to
provide the desired features.
If a schema name is given (for example,
CREATE AGGREGATE myschema.myagg ...
aggregate function is created in the speciﬁed schema. Otherwise it is created in the current schema.
An aggregate function is identiﬁed by its name and input data type(s). Two aggregates in the same
schema can have the same name if they operate on different input types. The name and input data
type(s) of an aggregate must also be distinct from the name and input data type(s) of every ordinary
function in the same schema. This behavior is identical to overloading of ordinary function names
(see CREATE FUNCTION).
Asimple aggregate function is made from one or two ordinary functions: a state transition function
,and an optional ﬁnal calculation function
.These are used as follows:
( internal-state, next-data-values ) ---> next-internal-state
( internal-state ) ---> aggregate-value
PostgreSQL creates a temporary variable of data type
to hold the current internal state of the
aggregate. At each input row, the aggregate argument value(s) are calculated and the state transition
functionis invoked withthe currentstate value and the newargumentvalue(s) tocalculatea new inter-
nal state value. After all the rows have been processed, the ﬁnal function is invoked once to calculate
the aggregate’s return value. If there is no ﬁnal function then the ending state value is returned as-is.
An aggregate function can provide an initial condition, that is, an initial value for the internal state
value. This is speciﬁed and stored in the database as a value of type
,but it must be a valid
external representation of a constant of the state value data type. If it is not supplied then the state
value starts out null.
If the state transition function is declared “strict”, then it cannot be calledwith null inputs. With such
atransition function, aggregate execution behaves as follows. Rows with any null input values are
ignored (the function is not called and the previous state value is retained). If the initial state value
is null, then at the ﬁrst row with all-nonnull input values, the ﬁrst argument value replaces the state
value, and the transition function is invoked at each subsequent row with all-nonnull input values.
This is handy for implementing aggregates like
.Note that this behavior is only available when
is the same as the ﬁrst
.When these types are different, you
must supply a nonnull initial condition or use a nonstrict transition function.
If the state transition functionis notstrict, then itwill be calledunconditionally ateach input row, and
mustdealwith null inputs and nullstate values for itself. This allows the aggregate author to have full
control over the aggregate’s handling of null values.
If the ﬁnal function is declared “strict”, then it will not be called when the ending state value is null;
instead a null resultwill be returned automatically. (Of course this is just thenormal behavior of strict
functions.) In any case the ﬁnal function has the option of returning a null value. For example, the
ﬁnal function for
returns null when it sees there were zero input rows.
Sometimes it is useful to declare the ﬁnal function as taking not just the state value, but extra pa-
rameters corresponding to the aggregate’s input values. The main reason for doing this is if the ﬁnal
function is polymorphic and the state value’s data type would be inadequate to pin down the result
type. These extra parameters are always passed as NULL (and so the ﬁnal function must not be strict
option is used), but nonetheless they are valid parameters. The ﬁnal
function could for example make use of
to identify the actual argument
type in the current call.
An aggregate can optionally support moving-aggregate mode, as described in Section 35.10.1. This
requires specifying the
parameters, and optionally the
parameters. Except for
rameters work like the corresponding simple-aggregate parameters without
;they deﬁne a separate
implementation of the aggregate that includes an inverse transition function.
The syntax with
inthe parameter list creates a special type of aggregate called an ordered-
set aggregate; or if
is speciﬁed, then a hypothetical-set aggregate is created. These
aggregates operate over groups of sorted values in order-dependent ways, so that speciﬁcation of an
input sort order is an essential part of a call. Also, they can have direct arguments, which are argu-
ments that are evaluated only once per aggregation rather than once per input row. Hypothetical-set
aggregates are asubclass of ordered-set aggregates inwhichsome of the directarguments are required
to match, in number anddata types, the aggregated argumentcolumns. This allows the values of those
direct arguments tobe added to the collection of aggregate-input rows as anadditional “hypothetical”
Aggregates that behave like
can sometimes be optimized by looking intoan indexinstead
of scanning every input row. If this aggregate can be so optimized, indicate it by specifying a sort
operator. The basic requirement is that the aggregate must yield the ﬁrst element in the sort ordering
induced by the operator; in other words:
SELECT agg(col) FROM tab;
must be equivalent to:
SELECT col FROM tab ORDER BY col USING sortop LIMIT 1;
Further assumptions are that the aggregate ignores null inputs, and that it delivers a null result if and
only if there were no non-null inputs. Ordinarily, a data type’s
operator is the proper sort operator
is the proper sort operator for
.Note that the optimizationwill never actually take
effect unless the speciﬁed operator is the “less than” or “greater than” strategy member of a B-tree
index operator class.
To be able to create anaggregate function, you must have
privilege on the argument types, the
state type(s), and the return type, as well as
privilege on the transition and ﬁnal functions.
The name (optionally schema-qualiﬁed) of the aggregate function to create.
The mode of an argument:
.(Aggregate functions do not support
ments.) If omitted, the default is
.Only the last argument can be marked
The name of an argument. This is currently only useful for documentation purposes. If omitted,
the argument has no name.
An input data type on which this aggregate function operates. To create a zero-argument aggre-
gate function, write
in place of the list of argument speciﬁcations. (An example of such an
In the old syntax for
,the input data type is speciﬁed by a
rameter rather than being written next to the aggregate name. Note that this syntax allows only
one input parameter. To deﬁne a zero-argument aggregate function with this syntax, specify the
). Ordered-set aggregates cannot be deﬁned with the old syntax.
The name of the state transition function to be called for each input row. For a normal
-argument aggregate function, the
+1 arguments, the ﬁrst being of type
and the rest matching the declared input data type(s) of the aggregate. The
function must return a value of type
.This function takes the current state
value and the current input data value(s), and returns the next state value.
For ordered-set(includinghypothetical-set) aggregates, thestatetransitionfunctionreceivesonly
the current state value and the aggregated arguments, not the direct arguments. Otherwise it is
The data type for the aggregate’s state value.
Theapproximateaverage size (inbytes) of the aggregate’s state value. If thisparameter is omitted
or is zero, a default estimate is used based on the
.The planner uses this
value to estimate the memory required for a grouped aggregate query. The planner will consider
using hash aggregation for such a query only if the hash table is estimated to ﬁt in work_mem;
therefore, large values of this parameter discourage use of hash aggregation.
The name of the ﬁnal function called to compute the aggregate’s result after all input rows
have been traversed. For a normal aggregate, this function must take a single argument of type
.The return data type of the aggregate is deﬁned as the return type of this
is not speciﬁed, then the ending state value is used as the aggregate’s result,
andthe return type is
For ordered-set (including hypothetical-set) aggregates, the ﬁnal function receives not only the
ﬁnal state value, but also the values of all the direct arguments.
is speciﬁed, then in addition to the ﬁnal state value and any direct argu-
ments, the ﬁnal function receives extra NULL values corresponding to the aggregate’s regular
(aggregated) arguments. This is mainly useful to allow correct resolution of the aggregate result
type when a polymorphic aggregate is being deﬁned.
The initial setting for the state value. This must be a string constant in the form accepted for the
.If not speciﬁed, the state value starts out null.
The name of the forward state transition function to be called for each input row in moving-
aggregate mode. This is exactly like the regular transition function, except thatits ﬁrst argument
andresult are of type
,which might be different from
The name of the inverse state transition function to be used in moving-aggregate mode. This
function has the same argument and result types as
,but it is used to remove a value
from the current aggregate state, rather than add a value to it. The inverse transition function
must have the same strictness attribute as the forwardstate transition function.
The data type for the aggregate’s state value, when using moving-aggregate mode.
The approximate average size (in bytes) of the aggregate’s state value, when using moving-
aggregate mode. This works the same as
The name of the ﬁnal function called to compute the aggregate’s result after all input rows have
been traversed, when using moving-aggregate mode. This works the same as
that its ﬁrst argument’s type is
and extra dummy arguments are spec-
iﬁed by writing
. The aggregate result type determined by
must match that determined by the aggregate’s regular implementation.
The initial setting for the state value, when using moving-aggregate mode. This works the same
The associated sort operator for a
-like aggregate. This is just an operator name
(possibly schema-qualiﬁed). The operator is assumed to have the same input data types as the
aggregate (which must be a single-argument normal aggregate).
For ordered-set aggregates only, this ﬂag speciﬁes that the aggregate arguments are to be pro-
cessed according to the requirements for hypothetical-set aggregates: that is, the last few di-
rect arguments must match the data types of the aggregated (
ﬂag has no effect on run-time behavior, only on parse-time resolution of the
data types and collations of the aggregate’s arguments.
The parameters of
can be written in any order, not just the order illustrated
In parameters that specify support function names, you can write a schema name if needed, for ex-
SFUNC = public.sum
.Do not write argument types there, however — the argument types of
the support functions are determined from other parameters.
If an aggregate supports moving-aggregate mode, it will improve calculation efﬁciency when the
aggregate is used as a window function for a window with moving frame start (that is, a frame start
mode other than
). Conceptually, the forward transition function adds input
values to the aggregate’s state when they enter the window frame from the bottom, and the inverse
transition function removes them again when they leave the frame at the top. So, when values are
removed, theyare always removedinthesameorder theywere added. Whenever the inverse transition
function is invoked, it will thus receive the earliest added but not yet removedargumentvalue(s). The
inverse transition function can assume that at least one row will remain in the current state after it
removes the oldest row. (When this would not be the case, the window function mechanism simply
starts a fresh aggregation, rather than using the inverse transition function.)
The forward transition function for moving-aggregate mode is not allowed to return NULL as the
new state value. If the inverse transition function returns NULL, this is taken as an indication that
the inverse function cannot reverse the state calculation for this particular input, and so the aggregate
calculation will be redone from scratch for the current frame starting position. This convention al-
lows moving-aggregate mode to be used in situations where there are some infrequent cases that are
impractical to reverse out of the running state value.
If no moving-aggregate implementation is supplied, the aggregate can still be used with moving
frames, butPostgreSQL will recompute the whole aggregation whenever the start of the framemoves.
Note that whether or not the aggregate supports moving-aggregate mode, PostgreSQL can handle a
moving frame end without recalculation; this is done by continuing to add new values to the aggre-
gate’s state. It is assumed that the ﬁnal function does not damage the aggregate’s state value, so that
the aggregationcan be continued even after an aggregate resultvalue has been obtained for one set of
The syntax for ordered-set aggregates allows
to be speciﬁed for both the last direct pa-
rameter and the last aggregated (
)parameter. However, the current implementation
restricts use of
in two ways. First, ordered-set aggregates can only use
not other variadic array types. Second, if the last direct parameter is
can be only one aggregated parameter and it must also be
.(In the representation
used in the system catalogs, these two parameters are merged into a single
cannot represent functions withmore thanone
parameter.) If the aggregate
is a hypothetical-set aggregate, the direct arguments that match the
the hypothetical ones; any preceding parameters represent additional direct arguments that are not
constrained to match the aggregated arguments.
Currently, ordered-set aggregates do not need to support moving-aggregate mode, since they cannot
be used as window functions.
See Section 35.10.
is a PostgreSQL language extension. The SQL standard does not provide for
user-deﬁned aggregate functions.
ALTER AGGREGATE, DROP AGGREGATE
Documents you may be interested
Documents you may be interested