Every date & time column supports extraction and bucketing transformation.
Bucketing adjusts or "rounds" the date to the chosen granularity (hour, day, week, month, quarter, year). Let's look at a few examples, consider entry `2021/01/10 10:00:00`, choosing
Bucketing adjusts or "rounds" the date to the chosen granularity (hour, day, week, month, quarter, year). Let's look at a few examples, consider entry `2021/01/10 10:00:00`, choosing
- hour yields `2021/01/10 10:00:00` (no change)
- day yields `2021/01/10` (hours, minutes and seconds get truncated)
- week yields `2021/01/04` (Monday of the same week)
- quarter yields `2021/01/01` (beginning of Q1 2021)
- year yields `2021/01/01` (first day of the year)
Extracting selects a desired part of the date and discards the rest. Again, using `2021/01/10 10:00:00` we can see how the common options transform the date to a single number:
- hour yields `10` (0-24)
- day of the week yields `7` (= Sunday, range from 1-7)
- day of the month yields `10` (1-31)
- day of the year yields `10` (1-365)
- week yields `1` (1-53)
- quarter yields `1` (1-4)
- year yields `2021`
it's also useful to define custom aliases for these values as needed. For example mapping `1` to `Monday` and so on.