If aggregation is not needed when querying timeseries, there is likely a more efficient way to retrieve data. By default, aggregate_timeseries() will always aggregate data, to avoid this, a user could pass a group_by=building_id and get individual timeseries for all buildings. For this scenario, building an Athena query that does not have any groups or aggregations would be quicker to retrieve (by about ~15% according to my tests).
Enhancement Request
If aggregation is not needed when querying timeseries, there is likely a more efficient way to retrieve data. By default,
aggregate_timeseries()
will always aggregate data, to avoid this, a user could pass a group_by=building_id
and get individual timeseries for all buildings. For this scenario, building an Athena query that does not have any groups or aggregations would be quicker to retrieve (by about ~15% according to my tests).