hadoop - 发票，颜色，类型，数量

我有一张桌子:

Invoice| Colour | Type

In_001 |  Red   | D
In_001 |  Red   | D
In_001 | Yellow | E
In_002 |  Red   | H

我希望得到的结果是整理出发票数量超过
购买了两件商品并显示数量。

见下文，预期结果:

Invoice | Colour | Type | Count

In_001  |  Red   |  D   | 2
In_001  | Yellow |  E   | 1

备注:由于仅购买了一项，因此不包含In_002。

我尝试了以下命令:

select invoice, colour, type, count(invoice) from t
group by invoice,colour,type
having count(invoice)>1;

结果是:

Invoice | Colour | Type | Count

In_001  |  Red   |  D   | 2

请帮忙。

最佳答案

分组依据是所有三个发票，颜色和类型。因此结果只会是这样。
我们可以尝试下面的查询来实现它:

select invoice,colour,type,count(invoice) from t where invoice in
(select invoice from t group by invoice having count(invoice) >1)
group by invoice,colour,type

关于hadoop - 发票，颜色，类型，数量，我们在Stack Overflow上找到一个类似的问题：https://stackoverflow.com/questions/33141592/