Get Counts, then Group By Values

I have a table in my Oracle SQL database as follows:

I'd like to count the number of times that a date shows up in a group, to get a result as follows:

How might I accomplish this as efficiently as possible?

You can unpivot then pivot:

create table date_table (
  date_a date, date_b DATE
);

insert into date_table (date_a, date_b) values (to_date('07-2025', 'MM-YYYY'), to_date('07-2025', 'MM-YYYY'));
insert into date_table (date_a, date_b) values (to_date('05-2025', 'MM-YYYY'), to_date('05-2025', 'MM-YYYY'));
insert into date_table (date_a, date_b) values (to_date('03-2025', 'MM-YYYY'), to_date('03-2025', 'MM-YYYY'));
insert into date_table (date_a, date_b) values (to_date('01-2025', 'MM-YYYY'), to_date('02-2025', 'MM-YYYY'));
insert into date_table (date_a, date_b) values (to_date('01-2025', 'MM-YYYY'), to_date('01-2025', 'MM-YYYY'));
insert into date_table (date_a, date_b) values (to_date('01-2025', 'MM-YYYY'), to_date('01-2025', 'MM-YYYY'));
insert into date_table (date_a, date_b) values (to_date('01-2025', 'MM-YYYY'), to_date('03-2025', 'MM-YYYY'));
insert into date_table (date_a, date_b) values (to_date('04-2025', 'MM-YYYY'), to_date('03-2025', 'MM-YYYY'));

select * from date_table
unpivot ( 
  dt for col in ( date_a, date_b )
)
pivot (
  count(*) for col in ( 
    'DATE_A' date_a, 
    'DATE_B' date_b 
  )
)
order  by dt desc;


DT                       DATE_A     DATE_B
-------------------- ---------- ----------
01-JUL-2025 00:00:00          1          1
01-MAY-2025 00:00:00          1          1
01-APR-2025 00:00:00          1          0
01-MAR-2025 00:00:00          1          3
01-FEB-2025 00:00:00          0          1
01-JAN-2025 00:00:00          4          2