如何将 3 个查询的结果合并到一张表中

Question

我有 3 个查询输出相同样式的结果。每个查询都遵循相同的结构。

查询中使用的表：

• registration_database：

包含注册参加调查但之前未注册参加调查的人的唯一 ID 号 [列：注册]（与 statefileid 连接）

• statefileid：

包含 [column: 'statefileid'] 的一大群人的唯一 ID 号，并包含有关他们最喜欢的颜色是什么的信息 [column: 'favcolor']

• 2016年数据：

包含 2016 年参加调查的每个人的 statefileid 值。有一个 [列：电子]，表示该人是通过电子方式“Y”还是亲自参加“N”调查（逻辑上，所有未通过电子方式参加调查的人必须亲自参加调查

• 2018年数据：

与 2016 年数据相同，但它包含 2018 年参加调查的每个人的信息

• 2020年数据：

与 2020 年数据相同，但它包含 2020 年参加调查的每个人的信息

public_scores_2020：包含一个 statefileid 列以连接到其他表和一个 [column: lrsupport] 给出从 1 到 100 的预测分数，无论它们可能支持左方向 (100) 还是支持右方向(1)

查询 1：

select
case when favcolor='r' then 'Red'
               when favcolor='b' then 'Blue'
               when favcolor='g' then 'Green'
               when favcolor='y' then 'Yellow'
               when favcolor='o' then 'Orange'
               when favcolor='u' then 'Unknown Fav Color'
               when favcolor='w' then 'White'
               else 'Unknown Fav Color' end as "Favorite_Color"
,round(sum((lrsupport)/100),0) as "left"
,round(sum((1-(lrsupport)/100)),0) as "right"
,count(registration) AS "new_registered"
FROM registration_database
left join
statefileid
on registration=statefileid
join 2016data AS a
using (statefileid)
left join public_scores_2020
using(statefileid)
where a.electronic = 'N'
group by 1
order by 1

查询 2：

select
case when favcolor='r' then 'Red'
               when favcolor='b' then 'Blue'
               when favcolor='g' then 'green'
               when favcolor='y' then 'yellow'
               when favcolor='o' then 'orange'
               when favcolor='u' then 'Unknown Fav Color'
               when favcolor='w' then 'White'
               else 'Unknown Fav Color' end as "Favorite_Color"
,round(sum((lrsupport)/100),0) as "left"
,round(sum((1-(lrsupport)/100)),0) as "right"
,count(registration) AS "new_registered"
FROM registration_database
left join
statefileid
on registration=statefileid
join 2018data AS a
using (statefileid)
left join public_scores_2020
using(statefileid)
where a.electronic = 'N'
group by 1
order by 1

查询 3：

select
case when favcolor='r' then 'Red'
               when favcolor='b' then 'Blue'
               when favcolor='g' then 'green'
               when favcolor='y' then 'yellow'
               when favcolor='o' then 'orange'
               when favcolor='u' then 'Unknown Fav Color'
               when favcolor='w' then 'White'
               else 'Unknown Fav Color' end as "Favorite_Color"
,round(sum((lrsupport)/100),0) as "left"
,round(sum((1-(lrsupport)/100)),0) as "right"
,count(registration) AS "new_registered"
FROM registration_database
left join
statefileid
on registration=statefileid
join 2020data AS a
using (statefileid)
left join public_scores_2020
using(statefileid)
where a.electronic = 'N'
group by 1
order by 1

每个表查询的输出如下所示，这是针对引用 2016data 的查询

我正在尝试组合 3 个查询的输出，以便我仍然可以获得按最喜欢的颜色行排序的结果 = 红色、蓝色、绿色等 9 列：2016left、2016right、2016new_registered、2018left、2018right、2018new_registered ,2020left, 2020right, 2020new_registered。

谁能帮我解决这个问题？

Answer 1

使用公用表表达式组合按年份拆分的表，这将降低查询的复杂性。这只是一个示例，如果不进一步努力，它可能无法正常工作：

with all_data as (
        select statefileid, lrsupport, electronic, 2016 as Year
        from 2016data
        union all
        select statefileid, lrsupport, electronic, 2018 as Year  
        from 2018data
        union all
        select statefileid, lrsupport, electronic, 2020 as Year
        from 2020data
        )

SELECT
  CASE
               when stf.favcolor='r' then 'Red'
               when stf.favcolor='b' then 'Blue'
               when stf.favcolor='g' then 'green'
               when stf.favcolor='y' then 'yellow'
               when stf.favcolor='o' then 'orange'
               when stf.favcolor='u' then 'Unknown Fav Color'
               when stf.favcolor='w' then 'White'
               else 'Unknown Fav Color'
  end as "Favorite_Color"
, count(case when a.year = 2016 then reg.registration end) AS "new_registered_2016"
, count(case when a.year = 2018 then reg.registration end) AS "new_registered_2018"
, count(case when a.year = 2020 then reg.registration end) AS "new_registered_2020"
FROM registration_database AS reg
INNER JOIN statefileid AS stf ON reg.registration = stf.statefileid
INNER JOIN all_data AS a ON stf.statefileid = a.statefileid
LEFT JOIN public_scores_2020 AS score ON stf.statefileid = score.statefileid
WHERE a.electronic = 'N'
GROUP BY 1
ORDER BY 1

在整个查询中引用列时也请使用表名或别名，我还建议使用更明确的连接来帮助解决这个问题。我认为对表 statefileid 的联接可能是内部联接