Changing format of mongoDB aggregation result [duplicate] - mongodb

Question

This question already has an answer here:
Promote subfields to top level in projection without listing all keys
(1 answer)
Closed 4 years ago.
Currently I am running this query to calculate averages and to return the data in a specific format:
db.metrics.aggregate([
{
$unwind:"$data"
},
{
$group:{
_id:"$data.configName",
avg:{
$avg:"$data.linesCount"
},
data:{
$last:"$data"
},
date:{
$last:"$date"
}
}
}
]).pretty()
On a collection which contains objects in this format:
{
"_id" : {
"date" : 1526569274000,
}
"date" : "20150220",
"data" : [
{
"configName" : "aaa",
"linesCount" : 500,
"insertedLinesCount" : 658,
}
],
"applicationName" : "loader"
}
Which returns this result:
{
"_id" : "aaa",
"avg" : 500,
"data" : {
"configName" : "aaa",
"linesCount" : 500,
"insertedLinesCount" : 658,
"succeeded" : true
},
"date" : "20150220"
}
The details are correct but I'd like to change the format. Is there any way to take what is in the data object and return it so that the final result is a list of 1-1 mappings, like so:
{
"_id" : "aaa",
"avg" : 500,
"configName" : "aaa",
"linesCount" : 500,
"insertedLinesCount" : 658,
"fileFormat" : "",
"date" : "20150220"
}

Answer 1

You need to use the $project stage at the end of the result
db.collection.aggregate([
{
$unwind: "$data"
},
{
$group: {
_id: "$data.configName",
avg: {
$avg: "$data.linesCount"
},
data: {
$last: "$data"
},
date: {
$last: "$date"
}
}
},
{
$project: {
configName: "$data.configName",
insertedLinesCount: "$data.insertedLinesCount",
linesCount: "$data.linesCount",
succeeded: "$data.succeeded",
_id: 1,
avg: 1,
date: 1
}
}
])
above query gives you the following result... check it here
[
{
"_id": "aaa",
"avg": 500,
"configName": "aaa",
"date": "20150220",
"insertedLinesCount": 658,
"linesCount": 500,
"succeeded": true
}
]

Answer 2

Add a $replaceRoot stage after $group
{
$replaceRoot: {
newRoot: {
_id: "$_id",
avg: "$avg",
configName: "$data.configName"
...
}
}
}