Question

我们说我有查询

#standardSQL
SELECT a.person, score, dte
FROM
(
    SELECT 'Aaron' person UNION ALL
    SELECT 'Baron' person UNION ALL
    SELECT 'Cindy' person
)a
LEFT JOIN
(
    SELECT 'Aaron' person, 10 score, 20 dte UNION ALL
    SELECT 'Baron' person, 20 score, 20 dte UNION ALL
    SELECT 'Aaron' person, 10 score, 21 dte UNION ALL
    SELECT 'Cindy' person, 10 score, 21 dte UNION ALL
    SELECT 'Baron' person, 20 score, 21 dte
)b
on a.person = b.person
order by dte

结果是

Row person  score   dte  
1   Aaron   10      20   
2   Baron   20      20   
3   Aaron   10      21   
4   Baron   20      21   
5   Cindy   10      21

现在，我是否还可以获得＆＃34; Cindy＆＃34;得分为0而dte = 20？

所以期望是：

Row person  score   dte  
1   Aaron   10      20   
2   Baron   20      20   
3   Cindy   0       20   
4   Aaron   10      21
5   Baron   20      21   
6   Cindy   10      21

使用左连接并不是必需的。基本上，我只有1个表，但只想为每个dte和person拥有0的数据。

Answer 1

以下是BigQuery Standard SQL

#standardSQL
WITH persons AS (
  SELECT 'Aaron' person UNION ALL
  SELECT 'Baron' person UNION ALL
  SELECT 'Cindy' person UNION ALL
  SELECT 'Mike' person
), scores AS (
  SELECT 'Aaron' person, 10 score, 20 dte UNION ALL
  SELECT 'Baron' person, 20 score, 20 dte UNION ALL
  SELECT 'Aaron' person, 10 score, 21 dte UNION ALL
  SELECT 'Cindy' person, 10 score, 21 dte UNION ALL
  SELECT 'Baron' person, 20 score, 21 dte
), dtes AS (
  SELECT DISTINCT dte 
  FROM scores
)
SELECT a.person, IFNULL(score, 0) score, d.dte
FROM persons a CROSS JOIN dtes d 
LEFT JOIN scores b 
ON a.person = b.person AND d.dte = b.dte
ORDER BY dte, person

结果如下

person   score    dte    
Aaron    10       20     
Baron    20       20     
Cindy    0        20     
Mike     0        20     
Aaron    10       21     
Baron    20       21     
Cindy    10       21     
Mike     0        21

query将null值替换为data bigquery

1 个答案: