附加两个表 - Union不在SQL中工作（Teradata）

Question

我是SQL的初学者。 我正在使用Teradata。 我们试图将数据汇总在一起，以比较实际值与预测值。

这是我的代码：

select 
'Actuals' as fc_version,
A.tra_fiscal_year as fiscal_year,
A.tra_fiscal_week as fiscal_week,
A.tra_allocation_category_code as category,
A.tra_grand_pricing_sales_channel as grandparent_channel,
D.fc_prod_id,
coalesce(
case
when A.tra_ticket_product_code like 'A%' then 'Total'
else null
end,

case
when A.tra_grand_pricing_sales_channel = 'INMKT' Then 'In Market'
else 'All Other'
end )
as fc_sales_channel,

sum(A.tra_ticket_quantity) as units_sold,

from ((pd_plan_forecast_db.adm_rev_detail A
left outer join
pd_plan_forecast_db.adm_rev_prod_code_to_fc_id B
on (A.tra_ticket_product_code = B.product_code))
left outer join
pd_plan_forecast_db.adm_rev_ticket_code_to_fc_id C
on (A.tra_ticket_code = C.ticket_code)
left outer join
pd_plan_forecast_db.adm_rev_fc_prod_info D
on (coalesce(B.fc_prod_id, C.fc_prod_id) = D.fc_prod_id))

group by 1, 2, 3, 4, 5, 6, 7

union

select fc_version, fiscal_year, fiscal_week, category, null as grandparent_channel, fc_prod_id, fc_sales_channel, sum(units_sold) as units_sold
from pd_plan_forecast_db.adm_rev_fc
where fiscal_year = 2017 and fiscal_week = 1
group by 1,2,3,4,5,6,7

如果将第一部分粘贴到联合之前，则它具有正确的结果。 同样是工会后的第二部分。

但是整个代码一起给出了错误：SELECT失败。 [3654]相应的选择列表表达式是不兼容的。

这是我正在使用的表的一般概念。

Answer 1

您必须在两个部分（顶部和底部）以及相同的数据类型中具有相同数量的列。

据我所知，顶部查询中有9列，底部查询中有8列。

UPD ：根据作者的修复在下面的查询中将null转换为空字符串。

Answer 2

您应该在两个选择中具有相同数量的列，并且（在选择中数字似乎为8）但是
在你的联盟中，你不应该在第二个选择中使用别名

select 
'Actuals' as fc_version,
A.tra_fiscal_year as fiscal_year,
A.tra_fiscal_week as fiscal_week,
A.tra_allocation_category_code as category,
A.tra_grand_pricing_sales_channel as grandparent_channel,
D.fc_prod_id,
coalesce(
case
when A.tra_ticket_product_code like 'A%' then 'Total'
else null
end,

case
when A.tra_grand_pricing_sales_channel = 'INMKT' Then 'In Market'
else 'All Other'
end )
as fc_sales_channel,

sum(A.tra_ticket_quantity) as units_sold,

from ((pd_plan_forecast_db.adm_rev_detail A
left outer join
pd_plan_forecast_db.adm_rev_prod_code_to_fc_id B
on (A.tra_ticket_product_code = B.product_code))
left outer join
pd_plan_forecast_db.adm_rev_ticket_code_to_fc_id C
on (A.tra_ticket_code = C.ticket_code)
left outer join
pd_plan_forecast_db.adm_rev_fc_prod_info D
on (coalesce(B.fc_prod_id, C.fc_prod_id) = D.fc_prod_id))

group by 1, 2, 3, 4, 5, 6, 7

union

select 
    fc_version
    , fiscal_year
    , fiscal_week
    , category
    , null 
    , fc_prod_id
    , fc_sales_channel
    , sum(units_sold) 
from pd_plan_forecast_db.adm_rev_fc
where fiscal_year = 2017 and fiscal_week = 1
group by 1,2,3,4,5,6,7