In PostgreSQL do ranking window functions heed the window frame or act on the entire partition?

I am learning window functions, primarily with this page of the docs. I am trying to categorize the window functions according to whether they heed window frames, or ignore them and act on the partition.

This page of the docs clearly states first_value(), last_value(), and nth_value(), as well as aggregate functions, only consider the window frame. However, the docs don't say as directly whether the ranking functions (i.e. rank(), dense_rank(), percent_rank(), and cume_dist()) heed or ignore window frames.

My educated guess is that these ranking functions, along with all other window functions listed in Table 9.65 (excluding first_value(), last_value(), and nth_value(), and aggregate window functions) ignore window frames.

1. Is this correct?
2. Also, does the concept of peer groups in rank() and dense_rank() offer some intuition as to why these functions ignore window frames?

"Ranking" window functions ignore the frame within the window, because the frame itself is defined by the rank resulting from the window's order by.

You skipped row_number(), which also belongs with other ranking functions, ignoring, or applying outside/above/before the frame.

Be sure to take a look at Chapter 4.2.8. SQL Syntax. Value Expressions: Window Function Calls.

Here are some examples to illustrate: _{demo at db<>fiddle}

select*,sum(x)over w1,sum(x)over w2,sum(x)over w3,sum(x)over w4
from t
window w1 as(order by y)
     , w2 as(order by y rows between unbounded preceding
                                 and current row)
     , w3 as(order by y rows between unbounded preceding
                                 and unbounded following)
     , w4 as(order by y rows between current row
                                 and unbounded following);

select*,first_value(x)over w1,first_value(x)over w2,first_value(x)over w3,first_value(x)over w4
       ,nth_value(x,2)over w1,nth_value(x,2)over w2,nth_value(x,2)over w3,nth_value(x,2)over w4
       ,last_value(x) over w1,last_value(x) over w2,last_value(x) over w3,last_value(x) over w4
from t
window w1 as(order by y)
     , w2 as(order by y rows between unbounded preceding
                                 and current row)
     , w3 as(order by y rows between unbounded preceding
                                 and unbounded following)
     , w4 as(order by y rows between current row
                                 and unbounded following);

---

No difference here:

select*,row_number()over w1,row_number()over w2,row_number()over w3,row_number()over w4
       ,dense_rank()over w1,dense_rank()over w2,dense_rank()over w3,dense_rank()over w4
       ,rank()      over w1,rank()      over w2,rank()      over w3,rank()      over w4
from t
window w1 as(order by y)
     , w2 as(order by y rows between unbounded preceding
                                 and current row
                             exclude current row)
     , w3 as(order by y rows between unbounded preceding
                                 and unbounded following
                             exclude current row)
     , w4 as(order by y rows between current row
                                 and unbounded following
                             exclude current row);

select*,lag(x,2)over w1,lag(x,2)over w2,lag(x,2)over w3,lag(x,2)over w4
       ,lead(x,2)over w1,lead(x,2)over w2,lead(x,2)over w3,lead(x,2)over w4
from t
window w1 as(order by y rows between 1 preceding
                                 and current row)
     , w2 as(order by y rows between unbounded preceding
                                 and current row
                             exclude current row)
     , w3 as(order by y rows between unbounded preceding
                                 and current row
                             exclude ties)
     , w4 as(order by y rows between 1 preceding
                                 and current row
                             exclude ties)

select*,cume_dist()over w1,cume_dist()over w2,cume_dist()over w3,cume_dist()over w4
from t
window w1 as(order by y)
     , w2 as(order by y rows between unbounded preceding
                                 and current row
                             exclude current row)
     , w3 as(order by y rows between unbounded preceding
                                 and unbounded following
                             exclude current row)
     , w4 as(order by y rows between current row
                                 and unbounded following
                             exclude current row);

select*,ntile(3)over w1,ntile(3)over w2,ntile(3)over w3,ntile(3)over w4
       ,ntile(5)over w1,ntile(5)over w2,ntile(5)over w3,ntile(5)over w4
       ,ntile(6)over w1,ntile(6)over w2,ntile(6)over w3,ntile(6)over w4
from t
window w1 as(order by y)
     , w2 as(order by y rows between unbounded preceding
                                 and current row
                             exclude current row)
     , w3 as(order by y rows between unbounded preceding
                                 and unbounded following
                             exclude current row)
     , w4 as(order by y rows between current row
                                 and unbounded following
                             exclude current row);