在 sqldf / R 中的情况

Question

如果我有DF

我尝试如下使用sqldf

 Test <- sqldf("                SELECT A,
                                    case when A = '1' and B = '1' then Col else NULL end as Test_1, 
                                    case when A = '1' and B = '2' then Col else NULL end as Test_2, 
                                    case when A = '1' and B = '3' then Col else NULL end as Test_3, 
                                    case when A = '2' and B = '1' then Col else NULL end as Test_4, 
                                     case when A = '2' and B = '2' then Col else NULL end as Test_5, 
                                      case when A = '2' and B = '3' then Col else NULL end as Test_6 

                                   FROM  DF
                                   group by A;")

但是，对于 A 的每个案例和其他列 NULL，它只有 returns 结果，而不是我想要的

A Test1 Test2 Test3 Test4 Test5 Test6
1   A     C     E    Null  NULL  NULL
2  NULL  NULL  NULL   D      B    F

每个A我只得到一个结果

A Test1 Test2 Test3 Test4 Test5 Test6
1   G    NULL  NULL  Null  NULL  NULL
2  NULL  NULL  NULL  NULL   B    NULL

我做错了什么，有没有办法获得我需要的格式？

Answer 1

在 SQL 中被称为条件聚合（通常用于旋转数据），因为@Gregor 评论只是运行像 MAX() 这样的聚合（甚至 MIN() 也可以工作) 围绕 CASE 语句：

SELECT A,
       MAX(CASE WHEN A = '1' AND B = '1' THEN Col ELSE NULL END) as Test_1, 
       MAX(CASE WHEN A = '1' AND B = '2' THEN Col ELSE NULL END) as Test_2, 
       MAX(CASE WHEN A = '1' AND B = '3' THEN Col ELSE NULL END) as Test_3, 
       MAX(CASE WHEN A = '2' AND B = '1' THEN Col ELSE NULL END) as Test_4, 
       MAX(CASE WHEN A = '2' AND B = '2' THEN Col ELSE NULL END) as Test_5, 
       MAX(CASE WHEN A = '2' AND B = '3' THEN Col ELSE NULL END) as Test_6
FROM DF
GROUP BY A

在 sqldf / R 中的情况

case when in sqldf / R

sql

sqlite

r

sqldf