I have some complex aggregation pipelines I need to execute, but my clients are using app services. Aggregation needs to be done by connecting to the data source as I understand, which executes the query in atlas (cloud). And it takes a good 2-3 seconds in my machine, to do one aggregation and get the result back.

I know that there are some aggregation operators in RQL which are not exactly as powerful as to the actual aggregation.

I want to do some grouping, counting, and count by date fields properties (data/ dayOfMonth, dayOfYear…and so on) …etc. Is there a way to do at least some of those through realm objects?

e.g: How am I supposed to do a simple counting by a field, like the following pipeline result, using realm sdk?

[
  {
    $group: {
      _id: "$state",
      count: { $sum: 1 }
    }
  }, ...
]

1. We are using app services with sync (but clients can go offline sometimes, for a short while), so I want to know if it is possible to run the aggregation on realm objects which would be faster than aggregation on the actual collection
2. Here’s a partial document from the “tasks” collection I have.

   {
     _id: ObjectId('66fbbfb826a136f8f67b422a'),
     on: ISODate('2024-10-01T09:24:08.719Z'),
     robot: 'peter',
     status: 'Completed',
   }
   

   I want to get some data for analytics from this collection. Say I want to count tasks by day of week, by status. I can run this aggregation on the collection, but it is very slow with the atlas cloud DB.

   [
     {
       $group: {
         _id: {
           status: "$status",
           dayOfWeek: { $dayOfWeek: "$on" }
         },
         count: { $sum: 1 }
       }
     }
   ]
   

   With realm objects, I can just run a “for realm.objects(Task)” loop over all the documents and calculate this, but I want to know if there is another, more realm/mongoDB way to do it.

I am using the nodejs sdk with typescript, if that helps.