MongoDB Aggregation PHP, Group by Days - php

Question

I have the db with the following format
Array
(
[_id] => MongoId Object
(
[$id] => 53f4bf0e8db0d31b0ba802df
)
[userSession] => 580929792589634763f964479eee8721
[pageEnteredDate] => 1408548587
[pageLeftDate] => 1408548622
[userName] => User 1
[userId] => 33657
[pageView] => monitoring patients
[pageActions] => []
[pageTag] => 1-3-16-131-315
[timeSpent] => 35
)
Array
(
[_id] => MongoId Object
(
[$id] => 53f3d7008db0d33e61cae841
)
[userSession] => e04e5081c9482654030bacf3c8c90b21
[pageEnteredDate] => 1408488536
[pageLeftDate] => 1408489216
[userName] => user 2
[userId] => 4278
[pageView] => Surgery Staff
[pageActions] => [["BUTTON","Comment",1408488701],["A","Discussion",1408488712]]
[pageTag] => 1-3-5-148
[timeSpent] => 680
)
Array
(
[_id] => MongoId Object
(
[$id] => 53f3d7008db0d33gj1cae841
)
[userSession] => e04e5081c9482654030bahjhc8c90b21
[pageEnteredDate] => 1408488536
[pageLeftDate] => 1408489216
[userName] => user 3
[userId] => 428
[pageView] => Surgery Staff
[pageActions] => [["BUTTON","Comment",1408488701],["A","Discussion",1408488712]]
[pageTag] => 1-3-5-148
[timeSpent] => 680
)
pageEnteredDate is the date that i want to use
I want to group the data by pagetag and day. I mean on one day i should get the same pageTag only once.
So, from those 3 arrays it should display only 2 because 2 have the same pageTag and are on the same day.
Thanks
***code used
$result = $this->collection->aggregate(
array(
array(
'$group' => array(
'_id'=> array( 'pageTag' => '$pageTag','day' => array('$subtract' => array('$pageEnteredDate', 86400))),
'timeSpent' => array( '$sum' => '$timeSpent' ),
'lastView' => array( '$max' => '$pageEnteredDate' )
)
),
array('$skip' => 0),
array('$limit' => 20)
)
);

Answer 1

Your "date" values appear just to be numbers derived from epoch timestamps ( without the milliseconds ). But all you really want to do here is apply $group via the aggregation framework, and a little date math to group by "day":
$result = $collection->aggregate(array(
array(
'$group' => array(
'_id' => array(
'pageTag' => 'pageTag',
'day' => array(
'$subtract' => array(
'$pageEnteredDate',
array('$mod' => array(
'$pageEnteredDate',
60 * 60 * 24
))
)
)
),
'timeSpent' => array( '$sum' => '$timeSpent' ),
'lastView' => array( '$max' => '$pageEnteredDate' )
)
)
));
That basically says to group on both the "pageTag" values and using the "pageEnteredDate" to apply the math that essentially rounds the timestamp to the current day, so all values within the same day are the same.
You haven't said exactly what it is that you want to "group" here, so the examples are given of applying a $sum to the "timeSpent" value and using $max to identify the last timestamp value recorded on that day.
You can use any of the "grouping operators" in this way to suit your needs