Add Group By support for Aggregate Functions
We're really happy about the new aggregate functions that recently were announced. But without Group By support they fall short and will leave many disappointed. Please add Group By support as well.
Johan Ohlin
shared this idea
Thank you everyone for your feedback. We are marking this as complete.
93 comments
-
Anonymous
commented
As I'm sure the Azure Cosmos team fully realizes, many (if not all) real-life scenarios require cross-partition queries. Since Group By does not support cross-partition queries it's quite unfair to mark this as complete.
Group By cannot be considered implemented if cross-partition queries are not supported.If you need this feature too, please upvote: https://feedback.azure.com/forums/263030-azure-cosmos-db/suggestions/40964188-group-by
-
satishyeeti
commented
Anyupdate on GroupBy support for .netSDK ?
-
Nidhi Singh
commented
Any Updates on Group BY clause support for JAVA SDK
-
Jan Fruseth
commented
Running the following:
var result = queryqueryable.Where(d => orgnrs.Contains(d.Orgnr)).GroupBy(ar => ar.Timestamp.Date).Select(a => new { Date = a.Key, OutstandingAmout = a.Sum(ar => ar.Outstanding) }).ToList();Got following exeption: "Method 'GroupBy' is not supported., Windows/10.0.18362 cosmos-netstandard-sdk/3.3.0. "
Someone who actual got this work?
-
Jan Fruseth
commented
Support for .Net SDK 3.3? Where can I download this?
-
Burdyugov, Anton
commented
Why does group by is not working for an non-partitioned collections?
SELECT top 5 c.questionName, count(1) as count FROM c
group by c.questionName
This simple query returns 0-0 results on portal in Data Explorer. Am I doing smth wrong? -
jasaram
commented
Thanks team...
-
Bryan
commented
I am also unable to utilize group by across partitions. Unfortunately, without cross-partition support, group by is largely useless to us. Are there any plans to support cross-partition group by queries?
EDIT - Looks like cross-partition group by works with the SDK only right now and will eventually be included in the portal.
-
Valerio Natangelo
commented
Just tried this querying from the portal:
https://docs.microsoft.com/en-gb/azure/cosmos-db/sql-query-group-by
and it fails for cross-partition queries:
Cross partition query only supports 'VALUE <AggreateFunc>' for aggregates
I have to say that after waiting one year this is very disappointing.
-
Anonymous
commented
What is awaiting moderator approval??
-
Anonymous
commented
It seems like it is now supported in SDK 3.3 and above
https://docs.microsoft.com/en-gb/azure/cosmos-db/sql-query-group-by -
Ryan Bowden
commented
This is now done: https://docs.microsoft.com/en-gb/azure/cosmos-db/sql-query-group-by
-
Jesse White
commented
Looks like this has been added:
https://docs.microsoft.com/en-gb/azure/cosmos-db/sql-query-group-byOnly available in .NET SDK 3.3
-
Anonymous
commented
Seems like it is now supported
https://docs.microsoft.com/en-gb/azure/cosmos-db/sql-query-group-by
-
Stuart Hallows
commented
Any update please?
-
Smit, Steven (CTN)
commented
Please let us know if this is going to ship by end of October? I need some clarity on this key feature.
-
Ste Wilson
commented
https://twitter.com/AzureCosmosDB/status/1166357072604274688
so we should expect this by the end of October?
Please can you update if your not going to hit this.
-
Philip Lupercio
commented
We moved to Couchbase because of this, too. My applications just could not wait any longer and I have stakeholders I need to satisfy. They are also getting support for SQL windowing functions + GROUP BY as well as other very useful analytics functions... CTEs, UDFs, cross-document join, multi-document ACID transactions, etc...
Perhaps will reevaluate Cosmos DB in the years to come.
-
Anonymous
commented
They say that it should be out in 6-8 weeks. Please don't move to Couchbase (yet)
https://twitter.com/AzureCosmosDB/status/1166357072604274688
-
Anonymous
commented
I totally agree with Nate. It's been quite long since the status is just "coming soon". At least provide a reasonable ETA so that we can plan accordingly.