Question

我想创建一个SELECT查询，该查询产生30个字段长和1个记录厚度的视图。如果我们正在计算标题，请记录2条记录。

相关较大查询的具体部分是：

WHERE CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE = 1

这会产生所需的结果 - 记录EVENT_TIMESTAMP和CONTRACT_EFFECTIVE_DATE之间的天数为1。

但是我这样做了30天。类似的东西：

WHERE CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE = 1:30

这可能吗？我可以剪切n次粘贴查询30次并将数字1更改为相应的值。但肯定有更好的方法吗？

结果看起来像：

1 day   2 day   3 day   4 day
10       11       8      14

Answer 1

那是另一个PIVOT查询，又有很多剪切和粘贴。

假设您要计算与每天范围匹配的行数：

SELECT 
   COUNT(CASE WHEN CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE = 1 THEN 1 END) AS "1 day",
   COUNT(CASE WHEN CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE = 2 THEN 1 END) AS "2 day",
   COUNT(CASE WHEN CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE = 3 THEN 1 END) AS "3 day",
   COUNT(CASE WHEN CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE = 4 THEN 1 END) AS "4 day",
...
FROM tab
WHERE CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE <= 30

您也可以将计算放在派生表中，但这只是看起来更简单：

SELECT 
   COUNT(CASE WHEN diff = 1 THEN 1 END) AS "1 day",
   COUNT(CASE WHEN diff = 2 THEN 1 END) AS "2 day",
   COUNT(CASE WHEN diff = 3 THEN 1 END) AS "3 day",
   COUNT(CASE WHEN diff = 4 THEN 1 END) AS "4 day"
FROM 
 (
   SELECT CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE AS diff
   FROM tab
   WHERE diff <= 30
 ) AS dt

Answer 2

你在找这个吗？

WHERE CAST(EVENT_TIMESTAMP AS DATE) - CONTRACT_EFFECTIVE_DATE BETWEEN 1 AND 30

我认为BETWEEN就是你想要的。

在Teradata中排序

2 个答案: