I have very long time-series. I would like to aggregate at month or even year granularity.
Via derived time-series:
Via api: https://www.palantir.com/docs/foundry/api/v2/ontologies-v2-resources/time-series-properties/stream-points/
How could I aggregate at month level ? year level ?
Since month and year are not fixed time periods I think they’re not supported for this card. We do have the periodic aggregate card which should allow you to choose month or year but the semantics are a bit different (the docs explain the difference between rolling aggregate and periodic aggregate).
You could also stack multiple operations (do a rolling aggregate on a weekly basis and then do a rolling aggregate on that series of every 4 points to get approximately monthly values or every 52 points to get approximately yearly values). Or just approximate a month as 30 days and a year as 365 days and do a daily aggregation for that number of days
