MongoDb Aggregation conditional count group by in nested array - mongodb

Question

I am having trouble with the below situation.
I would like to count 'modAttmpt.retrySess.meta.k.gradedActionsFeedback' with value 'Very Good!' and group by 'modAttmpt.modId'.
So far tried below with no luck:
db.myCollection.aggregate()
.unwind("$modAttmpt")
.unwind("$modAttmpt.retrySess")
.project(
{
'modAttmpt.modId' :1,
'modAttmpt.retrySess.dur': 1,
'modAttmpt.retrySess.modPassSt': 1,
'modAttmpt.retrySess.meta' : 1 ,
'totalPass' :
{
$cond: [
{ $eq: ["$modAttmpt.retrySess.modPassSt", 1] },
1, 0
]
},
'passWithGood' :
{
$cond: { if: { $eq: ["$modAttmpt.retrySess.meta.k", 'gradedActionsFeedback'] }, then: 1, else: 0 }
}
}
)
.group
(
{
_id: "$modAttmpt.modId",
'avg': {$avg: '$modAttmpt.retrySess.dur'},
'totalPass' :{ $sum: '$totalPass'},
'passWithGood' : { $sum: '$passWithGood'}
}
)
Sample Data:
{
"_id" : ObjectId("617c10a8e1136c7cf51ffd9d"),
"modAttmpt" : {
"modId" : "mod_home",
"retrySess" : {
"meta" : [
{
"k" : "characterSelected",
"v" : "Male"
},
{
"k" : "gradedActionsFeedback",
"v" : "Very Good!"
}
],
"dur" : 432,
"modPassSt" : 1
}
},
"totalPass" : 1,
"passWithGood" : 0
},
{
"_id" : ObjectId("617c10a8e1136c7cf51ffd94"),
"modAttmpt" : {
"modId" : "mod_home",
"retrySess" : {
"meta" : [
{
"k" : "characterSelected",
"v" : "Male"
},
{
"k" : "gradedActionsFeedback",
"v" : "Very Good!"
}
],
}
},
"totalPass" : 1,
"passWithGood" : 0
},
{
"_id" : ObjectId("617c10a8e1136c7cf51ffd92"),
"modAttmpt" : {
"modId" : "mod_clq",
"retrySess" : {
"meta" : [
{
"k" : "gradedActionsFeedback",
"v" : "Very Good!"
}
],
}
},
"totalPass" : 1,
"passWithGood" : 0
},
{
"_id" : ObjectId("617c10a8e1136c7cf51ffd91"),
"modAttmpt" : {
"modId" : "mod_home",
"retrySess" : {
"meta" : [
{
"k" : "gradedActionsFeedback",
"v" : "Good!"
}
],
}
},
"totalPass" : 1,
"passWithGood" : 0
},
{
"_id" : ObjectId("617c10a8e1136c7cf51ffd95"),
"modAttmpt" : {
"modId" : "mod_clq",
"retrySess" : {
"meta" : [
{
"k" : "gradedActionsFeedback",
"v" : "Very Good!"
}
],
}
},
"totalPass" : 1,
"passWithGood" : 0
}

Answer 1

You can make use of $reduce operator to calculate modAttmpt.retrySess.meta.k.gradedActionsFeedback with value Very Good!
Replace your $project stage with the one below:
db.collection.aggregate([
{
"$project": {
"modAttmpt.modId": 1,
"modAttmpt.retrySess.dur": 1,
"modAttmpt.retrySess.modPassSt": 1,
"modAttmpt.retrySess.meta": 1,
"totalPass": {
$cond: [
{
$eq: [
"$modAttmpt.retrySess.modPassSt",
1
]
},
1,
0
]
},
"passWithGood": {
$cond: {
if: {
$eq: [
"$modAttmpt.retrySess.meta.k",
"gradedActionsFeedback"
]
},
then: 1,
else: 0
}
},
"passWithGood2": {
"$reduce": {
"input": "$modAttmpt.retrySess.meta",
"initialValue": 0,
"in": {
"$add": [
{
"$cond": {
"if": {
"$eq": [
"$$this.v",
"Very Good!"
]
},
"then": 1,
"else": 0,
},
},
"$$value",
]
}
}
}
}
},
{
"$group": {
_id: "$modAttmpt.modId",
"avg": {
$avg: "$modAttmpt.retrySess.dur"
},
"totalPass": {
$sum: "$totalPass"
},
"passWithGood": {
$sum: "$passWithGood2"
}
},
},
])
Mongo Playground Sample Execution