窗口函数在聚合之后计算。这意味着SUM(COUNT(*)) OVER ()将对整个结果集的计数求和...

SELECT
    category,
    count(id) as unit_count,
    COUNT(*) / SUM(COUNT(*)) OVER () AS percentage_of_all_units
FROM sales
GROUP BY category
ORDER BY count(id) DESC;

下面是根据所提到的场景的示例表：

create table sales(id number(3),
category    varchar2(20))
/

insert all
into sales values('01','televisions')
into sales values('02','televisions')
into sales values('03','televisions')
into sales values('04','televisions')
into sales values('05','televisions')
into sales values('06','televisions')
into sales values('07','mobile phones')
into sales values('08','televisions')
into sales values('09','mobile phones')
into sales values('10','mobile phones')
into sales values('11','televisions')
into sales values('12','televisions')
select * from dual;
/
select * from sales;
/
-- Required Query --
select category,count(*) count,
(count(*) / (select count(*) from sales) * 100)||'%' percentage_of_total
from sales
group by category;

我没有您的表或数据，所以下面的示例基于Scott的示例模式。
emp表中的行数：

SQL> select count(*) from emp;

  COUNT(*)
----------
        14

每项工作的雇员人数：

SQL> select job,
  2    count(*) cnt
  3  from emp
  4  group by job
  5  order by cnt desc;

JOB              CNT
--------- ----------
CLERK              4
SALESMAN           4
MANAGER            3
ANALYST            2
PRESIDENT          1

组合（即您的要求）：

SQL> select job,
  2    count(*) cnt,
  3    round((count(*) / (select count(*) from emp) * 100), 0) pct
  4  from emp
  5  group by job
  6  order by cnt desc;

JOB              CNT        PCT
--------- ---------- ----------
CLERK              4         29
SALESMAN           4         29
MANAGER            3         21
ANALYST            2         14
PRESIDENT          1          7

SQL>

或者，如@Mat所建议的，第3行可能是

round((count(*) / sum(count(*)) over () * 100), 0) pct

适用于您的案例：

select category,
  count(*) as unit_count,
  round((count(*) / (select count(*) from sales) * 100), 0) percentage_of_total,
  --
  round((count(*) / sum(count(*)) over () * 100), 0) percentage_of_total_2
from sales
group by category
order by unit_count desc;