Aggregation and Find Operator

[Tamil Arasi]

Hi i have an use case where i have find all the matching records and apply the aggregation like count ,sum , group by etc . Can i use such scanerios where i have to use find and aggregate function together or should i find then get the cursor object and have to perform aggregation operation separetely 

[Alan Reyes]

You can use the $match stage of the aggregation pipeline as a find, then a $group stage to apply your aggregation operation.

[Wan Bachtiar]

Hi Tamil,

As mentioned by Alan, you can use $match pipeline operator to filter documents.

For example, the following selects documents to process, then pipes the results to the $group pipeline operator to generate count.

db.collection.aggregate( [
  { $match: { $or: [ { score: { $gt: 70, $lt: 90 } }, { views: { $gte: 1000 } } ] } },
  { $group: { _id: null, count: { $sum: 1 } } }
] );

See Aggregation pipeline operators for a list of available operators in MongoDB v3.2.

Also worth mentioning:

• Place the $match as early in the pipeline as possible, because $match limits the total number of documents in the pipeline and minimize the amount of processing down the pipe.
• If you place a $match at the very beginning of the pipeline, the query can take advantage of indexes.

If you still have any questions, can you provide the following:

• MongoDB version that you are using.
• Example documents.
• Example results that you are after.
• Aggregation pipeline that you have tried so far.

Regards,

Wan.

​

[Tamil Arasi]

I guess Match is equivalent to HAVING clause in Postgres But i want something like Select city,count(*) from Employee where division = 'Chennai' GROUP BY city.
Iam using Mongo DB version 3.2 . 

[Virendra Agarwal]

Hi Tamil,

Aggergation framework does the same job you are asking.

$ Match operation is equivalent to find

$ Project is equivalent to select field operation

$ Group is group by.

And aggrgation comes with all $sum / $avg etc.

Regards

Virendra Agarwal

[Alan Reyes]

Try with:

db.Employee.aggregate(
[
  {"$match": { "division": "Chennai" } },
  {"$group": { "_id": "$city", "count": {"$sum": 1} } }
]
)

$match filters only Employee in division "Chennai", then the resulting dataset is grouped by city, and for each grouped document the "count" attriobute of the result is increased by 1.