删除同名但值不同的行

Question

所以基本上我在 sql 中运行这段代码：

SELECT p.Nome as Nome_pub, a.Nome as Nome_area, COUNT(*) as total
FROM publicacao p, emprestimo e, area_tematica a
WHERE p.Id=e.Publicacao_Id and p.Area_Tematica_Id=a.Id and (Data_hora>='2021-01-01' AND Data_de_devolucao<='2021-06-31')
GROUP by p.Nome  
ORDER BY Nome_area;

我得到以下输出

Nome_pub	Nome_area	total
name1	a	1
name2	b	1
name3	c	1
name4	d	3
name5	d	2
name6	d	2
name7	e	1

但我想要一个输出，它基本上基于 word2 为我提供最大 word3：（在我的情况下，删除带有“name5”和“name6”的行

Nome_pub	Nome_area	total
name1	a	1
name2	b	1
name3	c	1
name4	d	3
name7	e	1

有什么办法吗？

提前致谢！！

Answer 1

如果ROW_NUMBER 在您的版本中可用。

SELECT Nome_pub, Nome_area, total
(
    SELECT 
      p.Nome as Nome_pub
    , a.Nome as Nome_area
    , COUNT(*) as total
    , ROW_NUMBER() OVER (PARTITION BY a.Nome ORDER BY COUNT(*) DESC) as RN
    FROM publicacao p
    JOIN emprestimo e
      ON p.Id = e.Publicacao_Id
    JOIN area_tematica a
      ON a.Id = p.Area_Tematica_Id
    WHERE (        Data_hora >= DATE '2021-01-01' AND 
           Data_de_devolucao <  DATE '2021-07-01')
    GROUP by p.Nome, a.Nome
) q
WHERE RN = 1 
ORDER BY Nome_area;