Teradata 从多行中获取变量的值

Question

我有以下不同日期的日期和金额变量。根据最少的日期，我每个 ID 只需要一行。

我拥有的数据：

 ID     Date         Code      Amt
 101    2/1/2018      A90       25
 101    2/2/2018      B90       15
 101    2/5/2018      A90       25
 102    2/1/2018      B90       10

我要的数据：

 ID     A90_Date     B90_Date    A90_Amt  B90_Amt
 101    2/1/2018      2/2/2018       25    15
 102                  2/1/2018             10

我的查询：

Select ID, 
    min(case when Code='A90' then Date else Null end) as A90_Date,
    min(case when Code='B90' then Date else Null end) as B90_Date,
    min(case when Code='A90' then Amt else Null end) as A90_Amt,
    min(case when Code='B90' then Amt else Null end) as B90_Amt
  from Table 
  group by 1

虽然这给了我日期和金额值，但我担心它们是否准确来自超过 1MM 的记录。有没有一种有效的方法可以使用 Teradata 中的 Qualify/Partition 语句提取这些数据？

提前感谢您的帮助。

Answer 1

如果相同 ID/code 的金额不同，您可能会得到错误的结果。那么你必须在聚合之前过滤最早的日期，例如

with cte as
 (
   Select ID, Code, Date, Amt
   from Table 
   qualify row_number() over (partition by ID, Code order by date) = 1
 )   
Select ID, 
    min(case when Code='A90' then Date end) as A90_Date,
    min(case when Code='B90' then Date end) as B90_Date,
    min(case when Code='A90' then Amt  end) as A90_Amt,
    min(case when Code='B90' then Amt  end) as B90_Amt
from cte 
group by 1

根据您的数据，这可能更有效：

Select DISTINCT ID, 
    first_value(case when Code='A90' then Date end) over (partition by ID order by date) as A90_Date,
    first_value(case when Code='B90' then Date end) over (partition by ID order by date) as B90_Date,
    first_value(case when Code='A90' then Amt  end) over (partition by ID order by date) as A90_Amt,
    first_value(case when Code='B90' then Amt  end) over (partition by ID order by date) as B90_Amt  
from Table 
group by 1

Teradata 从多行中获取变量的值

Teradata Getting value of a variable from multiple rows

case

min

teradata

partition