How to get sum of values for Male using Mongodb aggregate - mongodb

Question

I am new to mongoDB and having trouble getting the sum of values. I would like to get the sum of Male grouping by Date.
How can I achieve this? I tried the below query, but did not work. I am having trouble retrieving the value of Male.
db.sales.aggregate(
[
{
$group : {
_id: "$date",
TotalMaleValue: { $sum: "$follower_demographics.gender.value" }
}
}
]
)
My documents look like this:
{
"_id" : ObjectId("566dd67aef3ccf85743c4b10"),
"date" : "somedate",
"follower_demographics" : {
"gender" : [
{
"key" : "Male",
"value" : 480
},
{
"key" : "Female",
"value" : 1705
}
]
}
}
Any help is appreciated.

Answer 1

If you are using MongoDB-3.2 or newer then the best way to do this is to $project your documents and use the $filter operator to return an array with only those sub-documents where "key" is "Male". The next stage in the pipeline will then be the $unwind stage where you denormalize your array. The final stage is the $group stage where you use the $sum accumulator operator to calculate sum of "value".
Of course the $match stage in the beginning of the the pipeline let you select only those documents that match your criteria. This can reduce the size of documents to process in the next stage of the pipeline.
db.sales.aggregate([
{ '$match': {
'follower_demographics.gender': { '$elemMatch': {'key': 'Male' } }
}},
{
'$project': {
'date': 1,
'gender': {
'$filter': {
'input': '$follower_demographics.gender',
'as': 'g',
'cond': { '$eq': [ '$$g.key', 'Male' ] }
}
}
}
},
{ '$unwind': '$gender' },
{ '$group': {
'_id': '$date',
'TotalMaleValue': { '$sum': '$gender.value' }
}}
])
Prior to MongoDB 3.2 you need to use the little known $redact operator after $match to return an array with only those sub-documents where "key" is "Male".
db.sales.aggregate([
{ '$match': {
'follower_demographics.gender': { '$elemMatch': { 'key': 'Male' }}
}},
{ '$redact': {
'$cond': [
{ '$or': [ { '$eq': [ '$key', 'Male' ] }, { '$not': '$key' } ] },
'$$DESCEND',
'$$PRUNE'
]
}},
{ '$unwind': '$follower_demographics.gender' },
{ '$group': {
'_id': '$date',
'TotalMaleValue': { '$sum': '$follower_demographics.gender.value' }
}}
])

Answer 2

Note that, gender is an array. So you need to modify your query a bit to group by an object inside an array.
$unwind, the follower_demographics.gender field. This would now give you a one document each for an object inside the gender array.
$match, only those documents, that you are interested in, i.e those having the gender - value as Male.
Now, $group by the gender- key and get a $sum of the value field.
Sample code:
db.sales.aggregate([
{$match:{"follower_demographics.gender.key":"Male"}},
{$unwind:"$follower_demographics.gender"},
{$match:{"follower_demographics.gender.key":"Male"}},
{$group:{"_id":"$date",
"sum_of_males":{$sum:"$follower_demographics.gender.value"}}}
])