We've noticed placing INGORE NULLS
inside and outside the expressions for FIRST_VALUE and LAST_VALUE yields the same semantic result. According to the docs it should be inside the expression. Is this expected behaviour or should Redshift only accept IGNORE NULLS
inside the expression.
SELECT LAST_VALUE(venuename) IGNORE NULLS
OVER (
PARTITION BY venuestate
ORDER BY venueseats
ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW
) AS last_val
FROM venue
yields same result as
SELECT LAST_VALUE(venuename IGNORE NULLS)
OVER (
PARTITION BY venuestate
ORDER BY venueseats
ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW
) AS last_val
FROM venue