Mongodb group and sum two arrays of similar data from aggregate facet - arrays

Question

I am running queries that target different levels of social data and have got this working with some fairly large aggregate pipelines which then run inside a $facet.
The results of the $facet section come out with two arrays I then want to combine. eg.
{
"first": [
{
"_id": {
"name": "one",
"_id": "1"
},
"date": [
"2017-09-07T00:00:00.000Z"
],
"data": [
1000
]
},
{
"_id": {
"name": "two",
"_id": "2"
},
"date": [
"2017-09-07T00:00:00.000Z"
],
"data": [
2000
]
}
],
"second": [
{
"_id": {
"name": "1",
"_id": "one"
},
"date": [
"2017-09-07T00:00:00.000Z"
],
"data": [
1000
]
},
{
"_id": {
"name": "two",
"_id": "2"
},
"date": [
"2017-09-07T00:00:00.000Z"
],
"data": [
2000
]
}
]
}
I would like to group these results together and add up the totals to end up with
[ {
"_id": {
"name": "1",
"_id": "one"
},
"date": [
"2017-09-07T00:00:00.000Z"
],
"data": [
2000
]
},{
"_id": {
"name": "2",
"_id": "two"
},
"date": [
"2017-09-07T00:00:00.000Z"
],
"data": [
4000
]
}]
I have a query that groups the individual responses to get them to this level but is missing something that allows me to merge across the two arrays. I have experimented with unwinding the first and second array but haven't been able to find a solution that combines the data across both arrays.
Any help, tips or knowledge greatly appreciated.

Answer 1

Use $concatArrays to merge arrays from $facets output followed by $group with $arrayElemAt to pick the first element for $first and $sum accumulator.
[
{
"$project": {
"combine": {
"$concatArrays": [
"$first",
"$second"
]
}
}
},
{
"$unwind": "$combine"
},
{
"$group": {
"_id": "$combine._id",
"date": {
"$first": {
"$arrayElemAt": [
"$combine.date",
0
]
}
},
"data": {
"$sum": {
"$arrayElemAt": [
"$combine.data",
0
]
}
}
}
}
]