Here is the query,
db.getCollection('_ad.insight').aggregate([
{
$match:{
date: {
$lte: ISODate('2018-12-31'),
$gte: ISODate('2017-01-01')
}
}
},
{
$project:{
_id: 0,
year: {$year: "$date"},
month: {$month: "$date"},
day: {$dayOfMonth: "$date"},
date: 1,
clicks: 1,
reach: 1
}
},
{
$group:{
_id: {
month: "$month",
year: "$year"
},
clicks: {$sum: "$clicks"},
reach: {$sum: "$reach"},
date: {$addToSet: "$date"}
}
},
{
$project:{
_id: 0,
month: "$_id.month",
year: "$_id.year",
clicks: 1,
reach: 1,
date: 1
}
}
]);
And the response I am getting,
/* 1 */
{
"clicks" : 1096,
"reach" : 33288,
"date" : [
ISODate("2018-01-01T00:00:00.000+05:00"),
ISODate("2017-12-31T00:00:00.000+05:00"),
ISODate("2017-12-28T00:00:00.000+05:00"),
ISODate("2017-12-26T00:00:00.000+05:00"),
ISODate("2017-12-24T00:00:00.000+05:00"),
ISODate("2017-12-23T00:00:00.000+05:00"),
ISODate("2017-12-25T00:00:00.000+05:00"),
ISODate("2017-12-29T00:00:00.000+05:00"),
ISODate("2017-12-22T00:00:00.000+05:00"),
ISODate("2017-12-21T00:00:00.000+05:00"),
ISODate("2017-12-30T00:00:00.000+05:00"),
ISODate("2017-12-20T00:00:00.000+05:00"),
ISODate("2017-12-27T00:00:00.000+05:00")
],
"month" : 12,
"year" : 2017
},
/* 2 */
{
"clicks" : 1629,
"reach" : 98113,
"date" : [
ISODate("2018-01-05T00:00:00.000+05:00"),
ISODate("2018-01-04T00:00:00.000+05:00"),
ISODate("2018-01-03T00:00:00.000+05:00"),
ISODate("2018-01-07T00:00:00.000+05:00"),
ISODate("2018-01-08T00:00:00.000+05:00"),
ISODate("2018-01-02T00:00:00.000+05:00"),
ISODate("2018-01-06T00:00:00.000+05:00")
],
"month" : 1,
"year" : 2018
}
Sample Collection: Its a flat structure, contains around 400 fields, but i am showing only those which i am using in query.
{
"_id" : ObjectId("5akjbrd51f193455adtrf6fc"),
"clicks" : 5,
"reach" : 10
"date" : ISODate("2018-01-06T00:00:00.000+05:00"),
"post_engagement" : 127,
"post_reactions" : 1,
"post_shares" : 0,
"qualificationfailed" : 0,
"qualificationfailed_conversion_value" : 0
}
Desired Output:
/* 1 */
{
"clicks" : 1096,
"reach" : 33288,
"date" : [
ISODate("2018-01-01T00:00:00.000+05:00"),//this shouldn't be here
ISODate("2017-12-31T00:00:00.000+05:00"),
ISODate("2017-12-28T00:00:00.000+05:00"),
ISODate("2017-12-26T00:00:00.000+05:00"),
ISODate("2017-12-24T00:00:00.000+05:00"),
ISODate("2017-12-23T00:00:00.000+05:00"),
ISODate("2017-12-25T00:00:00.000+05:00"),
ISODate("2017-12-29T00:00:00.000+05:00"),
ISODate("2017-12-22T00:00:00.000+05:00"),
ISODate("2017-12-21T00:00:00.000+05:00"),
ISODate("2017-12-30T00:00:00.000+05:00"),
ISODate("2017-12-20T00:00:00.000+05:00"),
ISODate("2017-12-27T00:00:00.000+05:00")
],
"month" : 12,
"year" : 2017
},
/* 2 */
{
"clicks" : 1629,
"reach" : 98113,
"date" : [
// ISODate("2018-01-01T00:00:00.000+05:00") this should be in this group
ISODate("2018-01-05T00:00:00.000+05:00"),
ISODate("2018-01-04T00:00:00.000+05:00"),
ISODate("2018-01-03T00:00:00.000+05:00"),
ISODate("2018-01-07T00:00:00.000+05:00"),
ISODate("2018-01-08T00:00:00.000+05:00"),
ISODate("2018-01-02T00:00:00.000+05:00"),
ISODate("2018-01-06T00:00:00.000+05:00")
],
"month" : 1,
"year" : 2018
}
The issue is, ISODate("2018-01-01T00:00:00.000+05:00") as you can see in output document 1, in date array the above mentioned date is on first index. It shows the "month" : 12 and "year" : 2017 as i am grouping by month and year. So my concern is, ISODate("2018-01-01T00:00:00.000+05:00") should belong to group number 2, that is 2 output document but its showing up in group 1.
I don't know what i am doing wrong, as its a simple pipeline. Please help!!
Take note that ISODate("2018-01-01T00:00:00.000+05:00")
is in UTC + 5. This means that this entry has this date 2017-12-31T19:00:00 on UTC time.
Mongo is grouping the dates according to UTC.
You might want to check this post for dealing with different timezones How to aggregate by year-month-day on a different timezone