DB2-行数函数-需要输出中的第一个分区行

我尝试了下面的SQL，但它不起作用

问题：它没有给出第一个分区行

我正在添加表详细信息和SQL。这是针对DB2数据库的

Books Table:  
book_id  
rating 
title 
isbn 
publisher_id


WITH cte_books AS (
    SELECT
        publisher_id,
        ROW_NUMBER() OVER (
            PARTITION BY publisher_id 
            ORDER BY rating DESC
        ) row_num,
        book_id, 
        rating,
        title
    FROM 
        books
    WHERE 
        publisher_id IS NOT NULL
)
SELECT 
  * 
FROM 
    cte_books
WHERE 
    row_num >= 1 AND 
    row_num >= 2;

实际上是

行数>=2

 WHERE 
         row_num = 1

将返回第一行

编辑： 如果您希望为那些有多行的人提供所有行，这是一种方法：

WITH cte_books AS (
    SELECT
        publisher_id,
        ROW_NUMBER() OVER (
            PARTITION BY publisher_id 
            ORDER BY rating DESC
        ) row_num,
        book_id, 
        rating,
        title
    FROM 
        books
    WHERE 
        publisher_id IS NOT NULL
)

SELECT * 
  FROM cte_books t 
 WHERE exists (SELECT 1 
                 FROM cte_books 
                WHERE row_num > 1 
                  AND pubisher_id = t.publisher_id)

---

将

ROW_NUMBER（）

更改为

COUNT（1）

，删除

ORDER BY

，并在ROW_num>1的位置使用

。嗨，Mike，我需要输出至少有2个分区的发布者id记录。row_num=1对我不起作用。好的，谢谢你的澄清-我已经编辑了我的答案-检查是否可以修改此SQL以提供输出，其中row_num=1还将显示至少有两个分区的记录？我已将CTE添加到我的答案中-我不理解你的评论-它将为有分区的记录显示所有记录不止一个出版商叫迈克！这正是我想要的。谢谢！
WITH cte_books AS (
    SELECT
        publisher_id,
        ROW_NUMBER() OVER (
            PARTITION BY publisher_id 
            ORDER BY rating DESC
        ) row_num,
        book_id, 
        rating,
        title
    FROM 
        books
    WHERE 
        publisher_id IS NOT NULL
)

SELECT * 
  FROM cte_books t 
 WHERE exists (SELECT 1 
                 FROM cte_books 
                WHERE row_num > 1 
                  AND pubisher_id = t.publisher_id)