我想在 BigQuery 中將多條記錄合並為一條記錄

Question

我想在 BigQuery 中將多條記錄合並為一條記錄

pdt status            dt      
a   Inctive         2022/02/07
a   Inctive         2022/02/08
a   Active          2022/02/10  
a   Active          2022/02/11
a   Active          2022/02/15  
a   Active          2022/02/16
a   Active          2022/02/17
b   Active          2022/03/13  
b   Active          2022/03/14

Output

pdt status          start_dt    end_dt
a   Inactive        2022/02/07  2022/02/08
a   Active          2022/02/10  2022/02/11
a   Active          2022/02/15  2022/02/17
b   Active          2022/03/13  2022/03/14

我想知道如何在 bigquery 中對 function 期間進行規范化？

Answer 1

您可以考慮以下。

SELECT pdt, status, MIN(dt) AS start_dt, MAX(dt) AS end_dt FROM (
  SELECT *, UNIX_DATE(PARSE_DATE('%Y/%m/%d', dt)) - COUNT(1) OVER (PARTITION BY pdt ORDER BY dt) AS part
    FROM sample_data
) GROUP BY pdt, status, part
-- ORDER BY 1, 3;

查詢結果