distinct 和 Over 分區

Question

我正在處理 bigquery 中的查詢，但遇到以下問題。 我的查詢如下：

SELECT DISTINCT (concat(rut,created_at)),*,ROW_NUMBER () OVER (PARTITION BY rut ORDER BY created_at ASC) AS ranking 
FROM table 
WHERE DATE(created_at) <= "2020-12-03" 
order by rut,ranking

問題是“DISCINT”並沒有消除重復數據，而是如果我消除了OVER PARTITION並將查詢保留如下，如果它消除了重復數據。

SELECT DISTINCT (concat(rut,created_at)) ,*
FROM table 
WHERE DATE(created_at) <= "2020-12-03" 
order by rut

顯然，這是 disint 和 OVER PARTITION 如何工作的問題，但我找不到解決方案

Answer 1

你需要這樣做：

select * ,ROW_NUMBER () OVER (PARTITION BY rut ORDER BY created_at ASC) AS ranking  
from (
SELECT DISTINCT (concat(rut,created_at)),*
FROM table 
WHERE DATE(created_at) <= "2020-12-03" 
) tt
order by rut,ranking