강사님 group by에서 궁금한게 있습니다.

with
temp_01 as 
(
	select b.category_id, b.category_name as category_name 
	, percentile_disc(0.25) within group (order by unit_price) as qt_1
	, percentile_disc(0.5) within group (order by unit_price) as qt_2
	, percentile_disc(0.75) within group (order by unit_price) as qt_3
	, percentile_disc(1.0) within group (order by unit_price) as qt_4
from nw.products a
	join nw.categories b on a.category_id = b.category_id
group by b.category_id
)
select * from temp_01;

• 2번

  with
  temp_01 as 
  (
  	select a.category_id, max(b.category_name) as category_name 
  	, percentile_disc(0.25) within group (order by unit_price) as qt_1
  	, percentile_disc(0.5) within group (order by unit_price) as qt_2
  	, percentile_disc(0.75) within group (order by unit_price) as qt_3
  	, percentile_disc(1.0) within group (order by unit_price) as qt_4
  from nw.products a
  	join nw.categories b on a.category_id = b.category_id
  group by a.category_id
  )
  select * from temp_01;

  2번은 강사님이 작성하신 코드입니다.

• 1번처럼 작성하지 않으시고 2번으로 작성하신 이유가 따로 있으신가요 ?

• 인덱스가 중복되는게 있으면 효율이 안좋아서

• fk인 a.category를 사용하는게 더 좋아서 선택하신건지 아니면 1번이나 2번이나 둘다 큰 의미가 없는건지 궁금합니다.