Teradata 行到列（需要合并双倍行数）

Question

我有一个查询 运行 如下 -

SELECT 
item1,
item2,
item3,
CASE WHEN item4 = 'type A' THEN COUNT(DISTINCT field) END AS dcount_field_A,
CASE WHEN item4 = 'type B' THEN COUNT(DISTINCT field) END AS dcount_field_B,
FROM table
GROUP BY
item1
item2
item3
item4

我的输出是：

Item1	Item2	Item 3	dcount_field_B	dcount_field_A
Item1Value1	Item2Value1	Item3Value1	NULL	1234
Item1Value1	Item2Value1	Item3Value1	1234	NULL

但我需要合并这些行，例如：

Item1	Item2	Item 3	dcount_field_B	dcount_field_A
Item1Value1	Item2Value1	Item3Value1	1234	1234

其他尝试调整：

• 从组中删除项目 4 给我一个错误 "Selected 非聚合值必须是关联组的一部分。"
• 我的 teradata 版本没有 PIVOT 选项
• 我也试过
的子查询

    (SELECT
    COUNT(DISTINCT(CASE WHEN item4 = 'type A' THEN field END)) AS dcount_field_A,
    COUNT(DISTINCT(CASE WHEN item4 = 'type B' THEN field END)) AS dcount_field_B
    FROM table
    GROUP BY item4),

and I get the error "Too many expressions in the select list of a subquery."

Answer 1

您需要条件聚合。 count(distinct) 看起来像：

SELECT item1, item2, item3,
       COUNT(DISTINCT CASE WHEN item4 = 'type A' THEN field END) AS dcount_field_A,
       COUNT(DISTINCT CASE WHEN item4 = 'type B' THEN field END) AS dcount_field_B,
FROM table
GROUP BY item1, item2, item3;