Hugo Kornelis continues on a quest:
The Sequence Project operator computes values for the “ranking functions”: functions where the results depend on other rows in the result set, such as ROW_NUMBER, RANK, DENSE_RANK, and NTILE.
A Sequence Project can be considered as somewhat similar in function as Compute Scalar: both operators add new columns to the data based on expression. But Compute Scalar works on expressions other columns from the same row and constant values as input. Sequence Project computes expressions that are based on preceding rows in the data stream as their input.
Read on to learn more about what this operator does and how it works.