使用聚合框架可以对集合中的文档进行变换和组合。用多个构件创建一个管道（pipeline），用于对一连串的文档进行处理。这些构件包括:
筛选(filtering)
投射(projecting)
分组(grouping)
排序(sorting)
限制(limiting)
跳过(skipping)
在MongoDB中实际的集合框架，需要将这些操作传给aggregate函数，例如：
（1）将文档的name字段投射出来
> db.post.aggregate({$project:{"name":1}})
{ "_id" : ObjectId("54a530c3ff0df3732bac1681"), "name" : "joe" }
{ "_id" : ObjectId("54a530c3ff0df3732bac1680"), "name" : "joe" }
{ "_id" : ObjectId("54a9700e1b5afd45354fd086") }
{ "_id" : ObjectId("54a9701c1b5afd45354fd087") }
{ "_id" : ObjectId("54a970281b5afd45354fd088") }
{ "_id" : ObjectId("54a970351b5afd45354fd089") }
{ "_id" : ObjectId("54a970781b5afd45354fd08a") }
{ "_id" : ObjectId("54a970831b5afd45354fd08b") }
{ "_id" : ObjectId("54a970901b5afd45354fd08c") }
{ "_id" : ObjectId("54a9709c1b5afd45354fd08d") }
{ "_id" : ObjectId("54aa8a90652d8bdfa0566d34") }
{ "_id" : ObjectId("54aa97b894dcf31069b590ca") }
{ "_id" : ObjectId("54aa97d794dcf31069b590cb") }
{ "_id" : ObjectId("54aa97f294dcf31069b590cc") }
{ "_id" : ObjectId("54aff7f43bd1048e7b585e39") }
> db.post.aggregate({$project:{"name":1,"_id":0}})
{ "name" : "joe" }
{ "name" : "joe" }
可以看的出了find的用法类似；
（2）以name字段的值对其分组并计算其个数
> db.post.aggregate({$group:{"_id":"$name","count":{$sum:1}}})
{ "_id" : null, "count" : 13 }
{ "_id" : "joe", "count" : 2 }
>
（3）对结果集进行降序排列
> db.post.aggregate({$sort:{"id":-1}})
{ "_id" : ObjectId("54aa97d794dcf31069b590cb"), "id" : 13, "fruit" : [ "apple", "kumquat", "orange", "fruit01" ] }
{ "_id" : ObjectId("54aa97b894dcf31069b590ca"), "id" : 12, "fruit" : [ "apple", "banana", "peach" ] }
{ "_id" : ObjectId("54aa8a90652d8bdfa0566d34"), "id" : 11, "test10" : 11 }
{ "_id" : ObjectId("54a9709c1b5afd45354fd08d"), "id" : 10, "test10" : 10 }
{ "_id" : ObjectId("54a970901b5afd45354fd08c"), "id" : 9, "test9" : 9 }
{ "_id" : ObjectId("54a970831b5afd45354fd08b"), "id" : 8, "test8" : 8 }
{ "_id" : ObjectId("54a970781b5afd45354fd08a"), "id" : 7, "test7" : 7 }
{ "_id" : ObjectId("54a970351b5afd45354fd089"), "id" : 6, "test6" : 6 }
{ "_id" : ObjectId("54a970281b5afd45354fd088"), "id" : 5, "test5" : 5 }
{ "_id" : ObjectId("54a9701c1b5afd45354fd087"), "id" : 4, "test4" : 4 }
{ "_id" : ObjectId("54a9700e1b5afd45354fd086"), "id" : 3, "test3" : 3 }
{ "_id" : ObjectId("54a530c3ff0df3732bac1681"), "id" : 2, "name" : "joe", "age" : 30, "sex" : 1, "school" : "marry" }
{ "_id" : ObjectId("54a530c3ff0df3732bac1680"), "id" : 1, "name" : "joe", "age" : 30, "comments" : [ "test2", "test9", "test5" ], "sex" : 1, "school" : "marry" }
{ "_id" : ObjectId("54aa97f294dcf31069b590cc"), "push" : { "fruit" : "fruit01" } }
{ "_id" : ObjectId("54aff7f43bd1048e7b585e39"), "username" : "sid", "loc" : { "ip" : "1.2.3.4", "city" : "springfield", "state" : "ny" } }
>
（4）限制显示排序后的前5个结果
> db.post.aggregate({$sort:{"id":-1}},{$limit:5})
{ "_id" : ObjectId("54aa97d794dcf31069b590cb"), "id" : 13, "fruit" : [ "apple", "kumquat", "orange", "fruit01" ] }
{ "_id" : ObjectId("54aa97b894dcf31069b590ca"), "id" : 12, "fruit" : [ "apple", "banana", "peach" ] }
{ "_id" : ObjectId("54aa8a90652d8bdfa0566d34"), "id" : 11, "test10" : 11 }
{ "_id" : ObjectId("54a9709c1b5afd45354fd08d"), "id" : 10, "test10" : 10 }
{ "_id" : ObjectId("54a970901b5afd45354fd08c"), "id" : 9, "test9" : 9 }
>
聚合框架类似关系行数据库中的select，不会对集合进行写入操作，结果只会返回给客户端。