How to group date quarterly wise? - mongodb

Question

I have documents which contain a date and I'm wondering how to group them according to quarterly basis?
My schema is:
var ekgsanswermodel = new mongoose.Schema({
userId: {type: Schema.Types.ObjectId},
topicId : {type: Schema.Types.ObjectId},
ekgId : {type: Schema.Types.ObjectId},
answerSubmitted :{type: Number},
dateAttempted : { type: Date},
title : {type: String},
submissionSessionId : {type: String}
});
1st quarter contains months 1, 2, 3. 2nd quarter contains months 4, 5, 6 and so on up-to 4th quarter.
My final result should be:
"result" : [
{
_id: {
quater:
},
_id: {
quater:
},
_id: {
quater:
},
_id: {
quater:
}
}

Answer 1

You could make use of the $cond operator to check if:
The $month is <= 3, project a field named quarter with
value as "one".
The $month is <= 6, project a field named quarter with
value as "two".
The $month is <= 9, project a field named quarter with
value as "three".
else the value of the field quarter would be "fourth".
Then $group by the quarter field.
Code:
db.collection.aggregate([
{
$project: {
date: 1,
quarter: {
$cond: [
{ $lte: [{ $month: "$date" }, 3] },
"first",
{
$cond: [
{ $lte: [{ $month: "$date" }, 6] },
"second",
{
$cond: [{ $lte: [{ $month: "$date" }, 9] }, "third", "fourth"],
},
],
},
],
},
},
},
{ $group: { _id: { quarter: "$quarter" }, results: { $push: "$date" } } },
]);
Specific to your schema:
db.collection.aggregate([
{
$project: {
dateAttempted: 1,
userId: 1,
topicId: 1,
ekgId: 1,
title: 1,
quarter: {
$cond: [
{ $lte: [{ $month: "$dateAttempted" }, 3] },
"first",
{
$cond: [
{ $lte: [{ $month: "$dateAttempted" }, 6] },
"second",
{
$cond: [
{ $lte: [{ $month: "$dateAttempted" }, 9] },
"third",
"fourth",
],
},
],
},
],
},
},
},
{ $group: { _id: { quarter: "$quarter" }, results: { $push: "$$ROOT" } } },
]);

Answer 2

You could use following to group documents quarterly.
{
$project : {
dateAttempted : 1,
dateQuarter: {
$trunc : {$add: [{$divide: [{$subtract: [{$month:
"$dateAttempted"}, 1]}, 3]}, 1]}
}
}
}

Answer 3

Starting in Mongo 5, it's a perfect use case for the new $dateTrunc aggregation operator:
// { date: ISODate("2012-10-11") }
// { date: ISODate("2013-02-27") }
// { date: ISODate("2013-01-12") }
// { date: ISODate("2013-03-11") }
// { date: ISODate("2013-07-14") }
db.collection.aggregate([
{ $group: {
_id: { $dateTrunc: { date: "$date", unit: "quarter" } },
total: { $count: {} }
}}
])
// { _id: ISODate("2012-10-01"), total: 1 }
// { _id: ISODate("2013-01-01"), total: 3 }
// { _id: ISODate("2013-07-01"), total: 1 }
$dateTrunc truncates your dates at the beginning of their quarter (the truncation unit). It's kind of a modulo on dates per quarter.
Quarters in the output will be defined by their first day (Q3 2013 will be 2013-07-01). And you can always adapt it using $dateToString projection for instance.