CODEWITHSUNDEEP
mongodb
jwerre
jwerre

Reputation: 9584

How can I aggregate collection and group by field count

I have a collection of users that looks something like this:

{
  "_id": ObjectId("54380a817a4b612a38e87613"),
  "email": "[email protected]",
  "ogp": [BIG NESTED COLLECTION... {}, {}, {}]
  "created": ISODate("2012-02-28T23:10:07Z"),
  "o_id": ObjectId("5438096f7a4b612a38e445f4")
  "geo": {"country":"US", "city":"Seattle", "longitude": 123, "latitude":123}

}

I'd like to get all the users location and group them by country and total. Something like this:

[ {country:"US",total:250,000}, {country:"GB",total:150,000}, ... ]

Currently I'm just grabbing all of the documents and parsing it on the server:

db.users.find({'geo.country': {$ne: null},'geo.city': {$ne: null}}, {'geo.country':1}, function(err, doc) {
    var data;
    doc = _.groupBy(doc, function(par) { return par.geo.country; });
    data = [];
    return _.each(doc, function(item, key, obj) {
        return data.push([key, obj[key].length]);
    });
});

The problem with this is there are 600,000+ documents and the query takes about 1 minute to execute. Would the "aggregate" function would help speed up this query? If so how would I do it?

Upvotes: 2

Views: 176

Answers (1)

Abe Miessler
Abe Miessler

Reputation: 85036

This should do it:

db.myCollection.aggregate([
    {"$group": {_id: "$geo.country", count:{$sum:1}}}
])

Upvotes: 1

Related Questions