row_number() over skew partition by clolumns

[Loudongfeng]

The following SQL will run very slow due to skew values in skew_col column:

select row_number() over (partition by skew_col) from some_table;

Is there any way to optimize it?

Thanks