Segregate same columns into two columns - snowflake-cloud-data-platform

Question

I have a table with columns:
Id- unique id for each name
Name- client name
Date- account opening date
Sample table:
Id
Name
Date
101
a
8/7/2022
102
b
6/6/2022
101
a
16/8/2022
104
d
13/8/2022
105
e
23/4/2022
Query I am using:
Select id, name,
Max(date) over (partition by id) as max,
Min(date) over (partition by id) as min
From table
Output:
Id
Name
Date
Max
Min
101
a
8/7/22
16/8/22
8/7/22
102
b
6/6/22
6/6/22
6/6/22
101
a
16/8/22
16/8/22
8/7/22
104
d
13/8/22
13/8/22
13/8/22
105
e
23/4/22
23/4/22
23/4/22
The question is to divide the date column into maximum date and minimum date. But with that also where ever there is same date for that I need that date only in max column with null in min.
clear understanding of the question can be done from expected output.
Expected output:
Id
Name
Date
Max
Min
101
a
8/7/22
16/8/22
8/7/22
102
b
6/6/22
6/6/22
101
a
16/8/22
16/8/22
8/7/22
104
d
13/8/22
13/8/22
105
e
23/4/22
23/4/22
If there is help for this question, then do let me know.

Answer 1

You can use IFF function or Case Statement
;With SampleData As
(
Select Id, Name, To_Date(Date) AS Date
From
(
Values(101, 'a', '07/08/2022')
, (102, 'b', '06/06/2022')
, (101, 'a', '08/16/2022')
, (104, 'd', '08/13/2022')
, (105, 'e', '04/23/2022')
) AS Data(Id, Name, Date)
)
Select id
, name
, Max(date) over (partition by id) as max
, IFF(Min(date) over (partition by id) = Max(date) over (partition by id)
, NULL
, Min(date) over (partition by id)) as min
From SampleData
Please note that my date format is different than yours.

Answer 2

Mova's answer is correct and concise, it also can be expressed as nested select again with a CTE for data:
with data(Id, Name, Date) as (
select column1, column2, to_date(column3, 'dd/mm/yyyy') from values
(101, 'a', '8/7/2022'),
(102, 'b', '6/6/2022'),
(101, 'a', '16/8/2022'),
(104, 'd', '13/8/2022'),
(105, 'e', '23/4/2022')
)
as
select
id,
name,
date,
max_date,
iff(max_date != min_date, min_date, null) as min_date
from(
select
id
,name
,date
,max(date) over(partition by id, name) as max_date
,min(date) over(partition by id, name) as min_date
from data
)
ID
NAME
DATE
MAX_DATE
MIN_DATE
101
a
2022-07-08
2022-08-16
2022-07-08
102
b
2022-06-06
2022-06-06
null
101
a
2022-08-16
2022-08-16
2022-07-08
104
d
2022-08-13
2022-08-13
null
105
e
2022-04-23
2022-04-23
null
or a CASE can be used instead of the IFF
select
id,
name,
date,
max_date,
case when max_date != min_date then min_date end as min_date
from(
...
this shows we are trying to get a null value in some cases or the value, which is what NULLIF does, so this can be convert to one short line:
select
id
,name
,date
,max(date) over(partition by id, name) as max_date
,nullif(min(date) over(partition by id, name), max(date) over(partition by id, name)) as min_date
from data
and we can reuse the max_date, to make the small also, thus:
select
id
,name
,date
,max(date) over(partition by id, name) as max_date
,nullif(min(date) over(partition by id, name), max_date) as min_date
from data