在SQL中使用Pivot

Question

我很少在SQL中使用pivot，并且无法绕过它。我有查询，为医疗机构生成患者和保险的表格。我想返回一行，显示患者拥有的所有保险。

示例数据：

ID  Insurance
1840    Medicaid Of New York
1841    Affinity Health Plan
1841    Medicaid Of New York
1842    Fidelis Care
1842    Medicaid Of New York

目标是显示如下数据：

我理解pivot的基本功能，但没有太多时间来弄清楚如何实现上述结果。我没有取得任何接近工作结果的东西，这可能是缺乏经验或疲劳。任何帮助将不胜感激。

Answer 1

类似这样的事情

with data as (
select id,insurance ,row_number() over ( partition by id order by insurance) rowid 
)


select d.id , [1] as ins1,[2] ins2,[3]ins3,[4] ins4
from data d
pivot 
(max(insurance) 
for rowid in ([1],[2],[3],[4])
)pv

Answer 2

我更喜欢条件聚合：

SELECT  ID,
        MAX(CASE WHEN RN = 1 THEN INSURANCE ELSE '' END) AS INSURANCE1,
        MAX(CASE WHEN RN = 2 THEN INSURANCE ELSE '' END) AS INSURANCE2,
        MAX(CASE WHEN RN = 3 THEN INSURANCE ELSE '' END) AS INSURANCE3,
        MAX(CASE WHEN RN = 4 THEN INSURANCE ELSE '' END) AS INSURANCE4
FROM (
      SELECT  ID,
              INSURANCE,
              ROW_NUMBER() OVER (PARTITION BY ID ORDER BY INSURANCE) RN
      FROM YourTable
     ) A
GROUP BY ID

This可以很好地阅读条件聚合或交叉标签与PIVOT

Answer 3

如果您最多有两个值（如示例数据和结果所示），则可以使用min()和max()：

select id,
       min(insurance) as insurance1,
       (case when min(insurance) <> max(insurance) then max(insurance)
        end) as insurance2
from t
group by id;

Answer 4

这是您正在寻找的。这将根据需要平整您的记录。

SELECT 
    ID, Insurance1, Insurance2, Insurance3, Insurance4
FROM (  SELECT 
            ID, 
            Insurance,
            RN = 'Insurance' + 
            CAST(ROW_NUMBER() OVER(PARTITION BY ID ORDER BY Insurance) AS CHAR(1))
        FROM MyDB.MySchema.MyTable
) AS d
PIVOT(
    MAX(Insurance) 
        for RN 
            in (Insurance1, Insurance2, Insurance3, Insurance4)
) AS p;