How do I access a specific element after aggregation in mongodb? - mongodb

Question

After aggregation pipeline, I get a list of objects, but there is no way to retrieve the Nth object.
See:
http://docs.mongodb.org/manual/reference/operator/aggregation/group/#retrieve-distinct-values
The doc has an output like so:
{ "_id" : 1, "item" : "abc", "price" : 10, "quantity" : 2, "date" : ISODate("2014-03-01T08:00:00Z") }
{ "_id" : 2, "item" : "jkl", "price" : 20, "quantity" : 1, "date" : ISODate("2014-03-01T09:00:00Z") }
{ "_id" : 3, "item" : "xyz", "price" : 5, "quantity" : 10, "date" : ISODate("2014-03-15T09:00:00Z") }
{ "_id" : 4, "item" : "xyz", "price" : 5, "quantity" : 20, "date" : ISODate("2014-04-04T11:21:39.736Z") }
{ "_id" : 5, "item" : "abc", "price" : 10, "quantity" : 10, "date" : ISODate("2014-04-04T21:23:13.331Z") }
This is a group of objects, but it is not in a list so you can't do stuff like:
results[1] to get the second object. How are you supposed to interact with this group?

Answer 1

First of all, if you use the db.collection.distinct(<fieldname>) function, you can get distinct values of a field as an array:
> db.animals.distinct("call")
["moo", "baa", "woof", "meow", "quack"]
and then you can dereference the result since it's an array.
> calls = db.animals.distinct("call")
> calls[3]
"meow"
Aggregating for distinct values is free of the big limitation of db.collection.distinct() in that it returns a cursor over the distinct values instead of a big array, which means there's no 16MB BSON limit, and the distinct function can use indexes to cover the operation. So use the aggregation approach when you have a gazillion distinct values but otherwise use the distinct function. While you could call .toArray() on the cursor and get all the results in an array, if you have so many results that you couldn't use db.collection.distinct() then that is a bad idea. You should iterate through the cursor and pick out those values that you want and do stuff with them:
> var k = 0
> var calls = db.animals.aggregate(<pipeline for distinct animal calls>)
> while (calls.hasNext()) {
var call = calls.next()
k++
if (k == 96) doStuff(call)
}
You can insert a $skip stage in the pipeline to have the server skip right to the first result that you want, if you also include a $sort to fix an order that the results will be returned in. If you know you only want up to a certain amount, you can also use $limit and then the .toArray() approach may be viable again.

Answer 2

Before MongoDB 2.6 the results from the aggregate method were returned as a single document with the results placed within an array. This changed to accommodate larger result sets that would exceed the 16MB BSON limit that was restricted by this form. The cursor that is returned now is actually an option, but turned on by default in the shell.
Provided your results are not to large, there is actually a "helper" method for .toArray() which just does what you want by turning the cursor results into an array. It does cursor iteration under the hood, but basically just hides that from you:
var results = db.collection.aggregate(pipeline).toArray()
Then just access the element, it's an array so n-1:
var result = results[8];
Similar methods are available to most drivers, or otherwise in the driver do not include the "cursor" option.