(PECL mongo >=1.3.0)
MongoCollection::aggregate — Perform an aggregation using the aggregation framework
$pipeline
[, array $options
] )$op
[, array $op
[, array $...
]] )The MongoDB » aggregation framework provides a means to calculate aggregated values without having to use MapReduce. While MapReduce is powerful, it is often more difficult than necessary for many simple aggregation tasks, such as totaling or averaging field values.
This method accepts either a variable amount of pipeline operators, or a single array of operators constituting the pipeline.
pipelineAn array of pipeline operators.
optionsOptions for the aggregation command. Valid options include:
"allowDiskUse"
Allow aggregation stages to write to temporary files
"cursor"
Options controlling the creation of the cursor object. This option causes the command to return a result document suitable for constructing a MongoCommandCursor. If you need to use this option, you should consider using MongoCollection::aggregateCursor().
"explain"
Return information on the processing of the pipeline.
"maxTimeMS"
Specifies a cumulative time limit in milliseconds for processing the operation on the server (does not include idle time). If the operation is not completed by the server within the timeout period, a MongoExecutionTimeoutException will be thrown.
Or
opFirst pipeline operator.
opThe second pipeline operator.
...Additional pipeline operators.
The result of the aggregation as an array. The ok will be set to 1 on success, 0 on failure.
When an error occurs an array with the following keys will be returned:
| 版本 | 说明 |
|---|---|
| 1.5.0 |
Added optional options argument
|
Example #1 MongoCollection::aggregate() example
The following example aggregation operation pivots data to create a set of author names grouped by tags applied to an article. Call the aggregation framework by issuing the following command:
<?php
$m = new MongoClient("localhost");
$c = $m->selectDB("examples")->selectCollection("article");
$data = array (
'title' => 'this is my title',
'author' => 'bob',
'posted' => new MongoDate,
'pageViews' => 5,
'tags' => array ( 'fun', 'good', 'fun' ),
'comments' => array (
array (
'author' => 'joe',
'text' => 'this is cool',
),
array (
'author' => 'sam',
'text' => 'this is bad',
),
),
'other' =>array (
'foo' => 5,
),
);
$d = $c->insert($data, array("w" => 1));
$ops = array(
array(
'$project' => array(
"author" => 1,
"tags" => 1,
)
),
array('$unwind' => '$tags'),
array(
'$group' => array(
"_id" => array("tags" => '$tags'),
"authors" => array('$addToSet' => '$author'),
),
),
);
$results = $c->aggregate($ops);
var_dump($results);
?>
以上例程会输出：
array(2) {
["result"]=>
array(2) {
[0]=>
array(2) {
["_id"]=>
array(1) {
["tags"]=>
string(4) "good"
}
["authors"]=>
array(1) {
[0]=>
string(3) "bob"
}
}
[1]=>
array(2) {
["_id"]=>
array(1) {
["tags"]=>
string(3) "fun"
}
["authors"]=>
array(1) {
[0]=>
string(3) "bob"
}
}
}
["ok"]=>
float(1)
}
The following examples use the » zipcode data set. Use mongoimport to load this data set into your mongod instance.
Example #2 MongoCollection::aggregate() example
To return all states with a population greater than 10 million, use the following aggregation operation:
<?php
$m = new MongoClient("localhost");
$c = $m->selectDB("test")->selectCollection("zips");
$pipeline = array(
array(
'$group' => array(
'_id' => array('state' => '$state'),
'totalPop' => array('$sum' => '$pop')
)
),
array(
'$match' => array(
'totalPop' => array('$gte' => 10 * 1000 * 1000)
)
),
);
$out = $c->aggregate($pipeline);
var_dump($out);
?>
以上例程的输出类似于：
array(2) {
["result"]=>
array(7) {
[0]=>
array(2) {
["_id"]=>
string(2) "TX"
["totalPop"]=>
int(16986510)
}
[1]=>
array(2) {
["_id"]=>
string(2) "PA"
["totalPop"]=>
int(11881643)
}
[2]=>
array(2) {
["_id"]=>
string(2) "NY"
["totalPop"]=>
int(17990455)
}
[3]=>
array(2) {
["_id"]=>
string(2) "IL"
["totalPop"]=>
int(11430602)
}
[4]=>
array(2) {
["_id"]=>
string(2) "CA"
["totalPop"]=>
int(29760021)
}
[5]=>
array(2) {
["_id"]=>
string(2) "OH"
["totalPop"]=>
int(10847115)
}
[6]=>
array(2) {
["_id"]=>
string(2) "FL"
["totalPop"]=>
int(12937926)
}
}
["ok"]=>
float(1)
}
Example #3 MongoCollection::aggregate() example
To return the average populations for cities in each state, use the following aggregation operation:
<?php
$m = new MongoClient;
$c = $m->selectDB("test")->selectCollection("zips");
$out = $c->aggregate(
array(
'$group' => array(
'_id' => array('state' => '$state', 'city' => '$city' ),
'pop' => array('$sum' => '$pop' )
)
),
array(
'$group' => array(
'_id' => '$_id.state',
'avgCityPop' => array('$avg' => '$pop')
)
)
);
var_dump($out);
?>
以上例程的输出类似于：
array(2) {
["result"]=>
array(51) {
[0]=>
array(2) {
["_id"]=>
string(2) "DC"
["avgCityPop"]=>
float(303450)
}
[1]=>
array(2) {
["_id"]=>
string(2) "DE"
["avgCityPop"]=>
float(14481.913043478)
}
...
[49]=>
array(2) {
["_id"]=>
string(2) "WI"
["avgCityPop"]=>
float(7323.0074850299)
}
[50]=>
array(2) {
["_id"]=>
string(2) "WV"
["avgCityPop"]=>
float(2759.1953846154)
}
}
["ok"]=>
float(1)
}
Example #4 MongoCollection::aggregate() with command options
To return information on how the pipeline will be processed we use the explain command option:
<?php
$m = new MongoClient;
$c = $m->selectDB("test")->selectCollection("zips");
$pipeline = array(
array(
'$group' => array(
'_id' => '$state',
'totalPop' => array('$sum' => '$pop'),
),
),
array(
'$match' => array(
'totalPop' => array('$gte' => 10 * 1000 * 1000)
)
),
array(
'$sort' => array("totalPop" => -1),
),
);
$options = array("explain" => true);
$out = $c->aggregate($pipeline, $options);
var_dump($out);
?>
以上例程的输出类似于：
array(2) {
["stages"]=>
array(4) {
[0]=>
array(1) {
["$cursor"]=>
array(3) {
["query"]=>
array(0) {
}
["fields"]=>
array(3) {
["pop"]=>
int(1)
["state"]=>
int(1)
["_id"]=>
int(0)
}
["plan"]=>
array(4) {
["cursor"]=>
string(11) "BasicCursor"
["isMultiKey"]=>
bool(false)
["scanAndOrder"]=>
bool(false)
["allPlans"]=>
array(1) {
[0]=>
array(3) {
["cursor"]=>
string(11) "BasicCursor"
["isMultiKey"]=>
bool(false)
["scanAndOrder"]=>
bool(false)
}
}
}
}
}
[1]=>
array(1) {
["$group"]=>
array(2) {
["_id"]=>
string(6) "$state"
["totalPop"]=>
array(1) {
["$sum"]=>
string(4) "$pop"
}
}
}
[2]=>
array(1) {
["$match"]=>
array(1) {
["totalPop"]=>
array(1) {
["$gte"]=>
int(10000000)
}
}
}
[3]=>
array(1) {
["$sort"]=>
array(1) {
["sortKey"]=>
array(1) {
["totalPop"]=>
int(-1)
}
}
}
}
["ok"]=>
float(1)
}
COPY ONE COLLECTION TO ANOTHER COLLECTION IN SAME DATABASE
db.myoriginal.aggregate( [ [ $match: [] ], [ $out: "mycopy" ] ] )
a LOT faster than doing many inserts in a forEach loop.
< 2 seconds to copy 50,000 documents each a few KB.
12GB of data in 1-2 minutes on a i5 PC.
Best part : it's non-blocking!
Target can't be a capped collection.
ymmv