influxdata/docs-v2

Add example output for SQL window ranking functions

jstirnaman opened this issue · 0 comments

          Thanks for pointing it out! These three functions are very similar when the values are different, but have slightly different behaviors when identical values are present. (I just learned this today!)

Both the rank and dense_rank repeat the ranking for the identical values. rank skips ranks for identical values but dense_rank continues. For example,

ID       RANK      ROW_NUMBER      DENSE_RANK
-------- --------- --------------- ----------
1        1         1               1
1        1         2               1
1        1         3               1
2        4         4               2

I found a good example to show the difference between rank, row_number, and dense_rank:
https://stackoverflow.com/questions/7747327/sql-rank-versus-row-number

Originally posted by @appletreeisyellow in #5854 (comment)

ROW_NUMBER : Returns a unique number for each row starting with 1. For rows that have duplicate values,numbers are arbitarily assigned.

Rank : Assigns a unique number for each row starting with 1,except for rows that have duplicate values,in which case the same ranking is assigned and a gap appears in the sequence for each duplicate ranking.