Set fields based on sub document aggregates - mongodb

Question

In the collection below I am trying to calculate total / sold_total / sold_percent by aggregating the copies sub doc.
Similarly, I want to calculate grand_total / sold_grand_total / sold_grand_percent by aggregating the inventory sub document.
I prefer to do this during writes/updates or using a MongoDB function/job instead during 'reads' for efficiency.
I have tried a couple of aggregate pipelines but sub-array unwinding the copies array clears everything above it. Any help appreciated, thanks.
{
"_id" : "xyz",
"store" : "StoreB",
"grand_total" : 7,
"sold_grand_total" : 5,
"sold_grand_percent" : 72,
"inventory" : [
{"title" : "BookA", "total" : 4, "sold_total" : 3, "sold_percent" : 75,
"copies" : [
{"_id": 1, "condition": "new", "sold": 1 },
{"_id": 2,"condition": "new", "sold": 1 },
{"_id": 3,"condition": "new", "sold": 0 },
{"_id": 4,"condition": "new", "sold": 1 }
]
},
{"title" : "BookB", "total" : 1, "sold_total" : 1, "sold_percent" : 100,
"copies" : [
{"_id": 1, "condition": "new", "sold": 1 }
]
},
{"title" : "BookC", "total" : 2, "sold_total" : 1, "sold_percent" : 50,
"copies" : [
{"_id": 1, "condition": "new", "sold": 1 },
{"_id": 2,"condition": "new", "sold": 0 }
]
}
]
}

Answer 1

There are multiple ways of going this. I am not sure what your architecture is.
These are the 2 different aggregates:
This one gives "total" and "sold_total"
[
{
"$unwind" : "$inventory"
},
{
"$unwind" : "$inventory.copies"
},
{
"$group": {
"_id": "$inventory.title",
"total": {
"$sum": "$inventory.copies.sold"
},
"sold_total": {
"$sum": 1
}
}
}]
Other gives grand_total / sold_grand_total
[
{
"$unwind" : "$inventory"
},
{
"$unwind" : "$inventory.copies"
},
{
"$group": {
"_id": null,
"total": {
"$sum": "$inventory.copies.sold"
},
"count": {
"$sum": 1
}
}
}]
You can do both together, by getting the entire object from a group by operation and giving performing the other group by on it. basically, project and pipeline it.