group collection data by multiple fields mongodb - mongodb

Question

Collection Structure
Order = new Schema
index: { type: Number, unique: true }
number: Date
status: { type: String, enum: ['success', 'failure'] }
created_at: { type: Date, default: Date.now }
updated_at: { type: Date, default: Date.now }
I need help with a query that returns me an array of objects having data as success count and failure count grouped by date.
Ex-
orders = {
28-10-2016:{
success_count: 10,
failure_count: 10
},
29-10-2016: {
success_count: 10,
failure_count: 10
}
}

Answer 1

With the aggregation framework, the result will be slightly different from your "desired" output as instead of having hash keys, you get an array of objects with the _id key having a value that represents you group by field. For instance, instead of
{
"28-10-2016":{
"success_count": 10,
"failure_count": 10
},
"29-10-2016": {
"success_count": 10,
"failure_count": 10
}
}
you'd have a better structure like
[
{
"_id": "28-10-2016",
"success_count": 10,
"failure_count": 10
},
"_id": "29-10-2016",
"success_count": 10,
"failure_count": 10
}
]
Accomplishing the above result would require using the $cond operator in the $sum accumulator operator. The $cond operator will evaluate a logical condition based on its first argument (if) and then returns the second argument where the evaluation is true (then) or the third argument where false (else). This converts the true/false logic into 1 and 0 numerical values that feed into $sum respectively:
"success_count": {
"$sum": {
"$cond": [ { "$eq": [ "$status", "success" ] }, 1, 0 ]
}
}
As a resulting pipeline, one needs to run the aggregation operation which uses the $dateToString operator in the _id key expression for the $group pipeline:
Orders.aggregate([
{
"$group": {
"_id": {
"$dateToString": {
"format": "%Y-%m-%d",
"date": "$created_at"
}
},
"success_count": {
"$sum": {
"$cond": [ { "$eq": [ "$status", "success" ] }, 1, 0 ]
}
},
"failure_count": {
"$sum": {
"$cond": [ { "$eq": [ "$status", "failure" ] }, 1, 0 ]
}
}
}
}
], function (err, orders){
if (err) throw err;
console.log(orders);
})
However, there is a more flexible and better performant approach which executes much faster than the above, where the most efficient data structure for your aggregation result follows the schema for example:
orders = [
{
"_id": "28-10-2016",
"counts": [
{ "status": "success", "count": 10 },
{ "status": "failure", "count": 10 }
]
},
{
"_id": "29-10-2016",
"counts": [
{ "status": "success", "count": 10 },
{ "status": "failure", "count": 10 }
]
}
]
Then consider running an alternative pipeline as follows
Orders.aggregate([
{
"$group": {
"_id": {
"date": {
"$dateToString": {
"format": "%Y-%m-%d",
"date": "$created_at"
}
},
"status": { "$toLower": "$status" }
},
"count": { "$sum": 1 }
}
},
{
"$group": {
"_id": "$_id.date",
"counts": {
"$push": {
"status": "$_id.status",
"count": "$count"
}
}
}
}
], function (err, orders){
if (err) throw err;
console.log(orders);
})