This is in reference to this question.
This is my data set:
[
{
\"rating\": 4,
\"ceatedAt\": ISODate(\"2016-08-08T15:32:41.262+0000\")
},
{
For grouping on weekly basis, run the following pipeline which mainly uses the Date Aggregation Operators to extract the date parts:
db.collection.aggregate([
{
"$project": {
"createdAtWeek": { "$week": "$createdAt" },
"createdAtMonth": { "$month": "$createdAt" },
"rating": 1
}
},
{
"$group": {
"_id": "$createdAtWeek",
"average": { "$avg": "$rating" },
"month": { "$first": "$createdAtMonth" }
}
}
])
and for monthly aggregates, interchange the $group key to use the created month field:
db.collection.aggregate([
{
"$project": {
"createdAtWeek": { "$week": "$createdAt" },
"createdAtMonth": { "$month": "$createdAt" },
"rating": 1
}
},
{
"$group": {
"_id": "$createdAtMonth",
"average": { "$avg": "$rating" },
"week": { "$first": "$createdAtWeek" }
}
}
])