How to use aggregate for group by in mongodb - mongodb

Question

I have collection that contains documents with below schema.
Schema
{
"categoryId": "1234",
"sellerId": "2323",
"productId": "121",
"rating": 1
},
{
"categoryId": "1235",
"sellerId": "2323",
"productId": "122",
"rating": -1
},
{
"categoryId": "1234",
"sellerId": "2323",
"productId": "123",
"rating": -1
},
{
"categoryId": "1235",
"sellerId": "2323",
"productId": "124",
"rating": 1
},
{
"categoryId": "1234",
"sellerId": "2323",
"productId": "125",
"rating": 1
},
{
"categoryId": "1234",
"sellerId": "2325",
"productId": "125",
"rating": 1
}
The rating can have values 1 or -1. I want to find all documents grouped by categoryId and sum of the ratings.
Example Result:
{categoryId: 1234, positiveRatingCount: 2, negativeRatingCount: 1}
This is what I have done so far:
ratingsCollection.aggregate(
{
$match: {sellerId: "2323" }
},
{
$group: {
_id: "$categoryId",
count: { $sum: "rating" }
}
}
);
I get the following result. I am able to group by category but not able to figure out to get count of positive and negative ratings.
[
{
"_id": "1234",
"count": 3
},
{
"_id": "1235",
"count": 2
}
]

Answer 1

You need to use $sum with the conditions($cond) where the rating is $gt or $lt then 0
db.collection.aggregate([
{ "$match": { "sellerId": "2323" } },
{ "$group": {
"_id": "$categoryId",
"positiveRatingCount": {
"$sum": { "$cond": [{ "$gt": [ "$rating", 0 ] }, "$rating", 0 ] }
},
"negativeRatingCount": {
"$sum": { "$cond": [{ "$lt": [ "$rating", 0 ] }, "$rating", 0 ] }
}
}}
])
Output
[
{
"_id": "1235",
"negativeRatingCount": -1,
"positiveRatingCount": 1
},
{
"_id": "1234",
"negativeRatingCount": -2,
"positiveRatingCount": 3
}
]