Ranking the results in mysql (mysql equivalents for 'dense_rank()' or 'row_number()' functions in oracle)

Question

Below is the sample data:

dept_id  salary
10       10000
10       20000
10       20000
10       30000
20       50000
20       60000

I want to g

Answer 1

Dunno if it's faster but seems like a lot less typing...

SELECT x.*
     , IF(@prev = dept_id
                , IF(salary <> @psalary,@i:=@i+1,@i:=@i)
         ,@i:=1) rank
     , @psalary := salary
     , @prev := dept_id prev 
  FROM my_table x
     , (SELECT @i:=1,@prev:='',@psalary:='') vars 
 ORDER 
    BY dept_id,salary;