Repository navigation
arrow::compute::kernels::temporal should support nanoseconds #2995
Description
Is your feature request related to a problem or challenge? Please describe what you are trying to do.
arrow::compute::kernels::temporal should support nanoseconds in similar way it supports seconds
Describe the solution you'd like
Add functions to extracts the nanoseconds of a given temporal array as an array of integers
Describe alternatives you've considered
None
Additional context
None
Activity
comphead commentedon Nov 1, 2022
Please assign this ticket to me
waitingkuo commentedon Nov 1, 2022
@comphead thank you
do you plan to do millisecond only or microsecond/nanosecond as well in this pr?
comphead commentedon Nov 1, 2022
@waitingkuo for now we need milliseconds. but if the PR is relatively small, we can include micro/nano as well
waitingkuo commentedon Nov 1, 2022
@comphead
extracting second in postgresql contains microseconds
willy=# select date_part('second', now()); date_part ----------- 21.070954 (1 row)since the resolution for it's timestamp is microsecond
https://www.postgresql.org/docs/7.3/datatype-datetime.html
In datafusion, our default Timestamp's resolution is nanoseconds, so I think we might need that
comphead commentedon Nov 1, 2022
@waitingkuo I'm wondering would be that better to expose nanoseconds instead? Having that we can always convert nano to less precision data like millis, micros?
waitingkuo commentedon Nov 1, 2022
@comphead expose nanosecond sounds great
[-]arrow::compute::kernels::temporal should support milliseconds[/-][+]arrow::compute::kernels::temporal should support nanoseconds[/+]alamb commentedon Nov 11, 2022
label_issue.pyautomatically added labels {'arrow'} from #2996