Mongodb: Use record count in aggregation after a group

Question

I'm trying to store the total number of records input to the pipe so I can use the number in a later calculation. How do I grab the number of inputs, then unwind, then use the number of inputs later in my calcs?

I can get the number by doing this:

db.articles.aggregate([
  {
    $count: "totalArticles"
  }
]}

I can get the rest of the data I want by doing this:

db.articles.aggregate([
  {
    $unwind: "$concepts"
  },
  {
    $group: {
      _id:    "$concepts.text",
      count: {
        $sum: 1
      },
      average: {
        $avg: "$concepts.relevance"
      },
    }
  }
])

What I'd really like to do is this:

db.articles.aggregate([
  {
    $count: "totalArticles"
  },
  {
    $unwind: "$concepts"
  },
  {
    $group: {
      _id:    "$concepts.text",
      count: {
        $sum: 1
      },
      average: {
        $avg: "$concepts.relevance"
      }
    }
  },
  {
    $project: {
      count: "$count",
      percent: {
        $divide: [ "$count", "$totalArticles" ]
      }
    }
  },
  {
    $sort: {
      count: -1
    }
  }
])

s7vr · Accepted Answer · 2018-01-16 20:14:59Z

You can use below aggregation query.

Initial $group to calculate the total count while $push the concepts field into array field. $$ROOT to access the whole doc.

Retain the total articles count in next $group.

Rest all stays the way you've.

db.articles.aggregate([
  {"$group":{
    "_id":null,
    "totalArticles":{"$sum":1},
    "concepts":{"$push":"$$ROOT.concepts"}
  }},
  {"$unwind":"$concepts"},
  {"$group":{
    "_id":"$concepts.text",
    "totalArticles":{"$first":"$totalArticles"},
    "count":{"$sum":1},
    "average":{"$avg":"$concepts.relevance"}
  }},
  {"$project":{
      "count": "$count",
      "percent": {
        "$divide": [ "$count", "$totalArticles" ]
      }
    }
  },
  {"$sort": {"count": -1}}
])

$facets is also an option where you can two queries in two separate pipeline followed by merge to continue with rest of stages.

That works well thank you! I've read about $$ROOT and understand there is a total memory size. Is there a way to reduce that by selecting fields at that stage with $project? As you can see I'm only using the concepts field in the rest of the calc.

Collectives™ on Stack Overflow

Mongodb: Use record count in aggregation after a group

1 Answer 1

1 Comment

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Linked

Related