Aggregations: Histogram Aggregation bounds offsets

[rnonnon-ebiz]

Hi,

I'm working with elasticsearch and 1st, it's an amazing tool, 2nd I would suggest something about histogram aggregation with extended bounds.
Extended bounds is a good idea but it could be fine if we can force the bounds with, say, a boolean or another method...

For instance:
I have these data : [100,101,102,103,104]
I'd like to create an histogram with an interval of 3.

Elasticsearch will produce :
[ [99,doc_count:2] , [102,doc_count:3] ]
The problem is first tick always start from 0 but if I want to start my bucket from 100, I can't...

Then I can't obtain :
[ [100,doc_count:3] , [103,doc_count:2] ]

(The only solution is to perform a script on each data which will shift my data in the right bucket... then : creepy performances)

[rnonnon-ebiz]

no answer???

[jpountz]

I think this feature makes sense. Date histograms have ways to move the origin for timezone-related reasons so I think normal histograms should have that feature too.

[rnonnon-ebiz]

Great !
Thanks for your answer ;)

[zazasa]

Hi, i need this feature too.