Question

我有一张包含以下数据的表格：

id, foreign_id, contact, asset

对于具有不同foreign_id的同一联系人，有许多条目，我想要做的是获取基于asset的{{1}}值。

因此，让我们假设我们有一组这样的数据：

max(id)

我想写一个返回此内容的查询：

id: 1, foreign_id: 15, contact: email@email.com, asset: 1
id: 2, foreign_id: 15, contact: email@email.com, asset: 2
id: 2, foreign_id: 15, contact: email@email.com, asset: 2
id: 3, foreign_id: 14, contact: email@email.com, asset: 3
id: 4, foreign_id: 14, contact: emailTwo@email.com, asset: 1
id: 5, foreign_id: 15, contact: emailTwo@email.com, asset: 5
id: 6, foreign_id: 15, contact: emailTwo@email.com, asset: 12

如您所见，它为每个foreign_id的每个联系人获取最大资产。然后我还希望每行的id: 2, foreign_id: 15, contact: email@email.com, asset: 2, share_number: 2 id: 3, foreign_id: 14, contact: email@email.com, asset: 3, share_number: 1 id: 4, foreign_id: 14, contact: emailTwo@email.com, asset: 1, share_number 1 id: 5, foreign_id: 15, contact: emailTwo@email.com, asset: 12, share_number: 2与foreign_id排序的特定联系人有关。

这是在Marketing Cloud的SQL Server版本中编写的。我不能使用任何声明或变量。

Answer 1

一个选项使用ROW_NUMBER：

SELECT id, foreign_id, contact, asset,
    ROW_NUMBER() OVER (PARTITION BY contact ORDER BY id) rn
FROM
(
    SELECT id, foreign_id, contact, asset,
        ROW_NUMBER() OVER (PARTITION BY foreign_id, contact ORDER BY asset DESC) rn
    FROM yourTable
) t
WHERE t.rn = 1;

您的预期输出似乎没有任何特定顺序，但您可以通过在我的查询中添加GROUP BY子句来轻松强加一个。

修改：我在外部查询中添加了对ROW_NUMBER的另一个调用，该调用对共享同一联系人的记录进行编号。我假设排序由id列给出，但我不确定。将ORDER BY id替换为您要使用的列。

根据另外两列选择最多一列

1 个答案: