The problem, it's for example, in level 2, he count 2 value. But (watch the first picture) ID 120555 is level 1 and more level 2 if we refere to its date. So for level 2 he must count 1 and not 2 value !
I know somes similars topic was created, but i haven't found any cases behaving like mine, or maybe I misunderstand the explanations ?
I don't think you can run that type of aggregation in Lens. You can aggregate by id and then show the latest value of level ordered by timestamp which is not exactly what you want.
The only option I see is to use a Latest Transform to keep a separate index with just the latest document per id and visualize that.
I tried with several transforms by modifying the duration(100d / 7d / 1d). But it's the same result. Documents of new index pattern was deleted one by one after 15min. It's maybe because of another setting. But I'm having trouble seeing what
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.