ROW_ NUMBER()、RANK()、DENSE_ Rank difference

ROW_NUMBER()、RANK()、DENSE_RANK difference

ROW_NUMBER()： Sort , There will be no duplicate sorting values . For two equal numbers , Sequence number is inconsistent

The number	Sort sequence number
1	1
1	2
2	3

select
  id,
  row_number()
  over(order by id) rn
from 
  data

DENSE_RANK()： Sort , There may be duplicate values . For two equal numbers , The sequence number is consistent

The number	Sort sequence number
1	1
1	1
2	2

select
  id,
  dense_rank()
  over(order by id) rn
from 
  data

RANK()： Sort , There may be duplicate values . For two equal numbers , The sequence number is consistent , But the total will decrease

The number	Sort sequence number
1	1
1	1
2	3

select
  id,
  rank()
  over(order by id) rn
from 
  data

If in the window over() No sort field , So what's the result ？

select
  id,
  rank()
  over()
from 
  data

The number	Sort sequence number
1	1
1	1
2	1

If in the window over() No sort field , It uses row_number() Sort , What is the result? ？

select
  id,
  row_number()
  over()
from 
  data

The number	Sort sequence number
1	1
2	2
1	3

The reason for this phenomenon ：

stay over() Window , All numerical orders are the same . about rank Because repetition is allowed , So it's all 1, and row_number() There will be no duplicate sort sequence number , So it will 1,2,3,4, etc. , But the result of each sorting is different .