SQL Server SQL Query中的Excel四分位函数

Question

我有一组这样的值：

40
50
50
66
83
100
100
100
100
100
100
100
100
100
100
100
100

当我在 excel 中执行四分位数函数时，我得到了第一个四分位数 (25)、第二个四分位数 (50)、第三个四分位数 (75) 和最大值 (100) 的这四个值

四分位数 = 83.33, 100, 100, 100

所以当我比较一个获得 80% 的销售代表时，根据 excel 计算，他们将落在最后四分之一。

我需要在 sql 中重做相同的功能，并在下面给出了我的代码。

declare @sales table(
salesRepId int,
percentageSales int)

insert into @sales(salesRepId, percentageSales)
values(1,40)
,(2,50)
,(3,50)
,(4,66.7)
,(5,83.33)
,(6,100)
,(7,100)
,(8,100)
,(9,100)
,(10,100)
,(11,100)
,(12,100)
,(13,100)
,(14,100)
,(15,100)
,(16,100)
,(17,100);

with quintile as(
            select percentagesales, ntile(4) over(order by percentagesales) 
as quintile
            from (select distinct percentagesales from @sales) as s
            )
            select salesrepid, r.percentagesales, q.quintile
            from @sales r
            join quintile q on r.percentagesales = q.percentagesales
            order by q.quintile, percentagesales

当我运行它时，我得到以下结果集： Query results

salesrepid  percentagesales quintile
1   40  1
2   50  1
3   50  1
4   66  2
5   83  3
6   100 4
7   100 4
8   100 4
9   100 4
10  100 4
11  100 4
12  100 4
13  100 4
14  100 4
15  100 4
16  100 4
17  100 4

根据sql，80%会落在中间四分位。

SQL查询中如何获取类似于excel的四个百分位数

Answer 1

我尝试了 percentile_cont 建议的 Larnu，我得到四个四分位数的值

83,100,100,100

但从技术上讲，它应该是 83.33,100,100,100

这是我目前的查询

declare @sales table(
salesRepId int,
percentageSales int)

insert into @sales(salesRepId, percentageSales)
values(1,40.00)
,(2,50.00)
,(3,50.00)
,(4,66.77)
,(5,83.33)
,(6,100.00)
,(7,100.00)
,(8,100.00)
,(9,100.00)
,(10,100.00)
,(11,100.00)
,(12,100.00)
,(13,100.00)
,(14,100.00)
,(15,100.00)
,(16,100.00)
,(17,100.00);

with quintile as(
            select percentagesales, ntile(100) over(order by percentagesales) as quintile
            from (select distinct percentagesales from @sales) as s
            )
            select salesrepid, r.percentagesales, q.quintile
            from @sales r
            join quintile q on r.percentagesales = q.percentagesales
            order by q.quintile, percentagesales;

declare @p Decimal(2,2) = 0.25;
    with quartile as(
            select 
            salesRepId,
            percentageSales 
            ,cast(percentile_cont(.25)  within group(order by percentagesales) over() as decimal(36,2)) as quartile25th
            ,cast(percentile_cont(.5)  within group(order by percentagesales) over() as decimal(36,2)) as quartile50th
            ,cast(percentile_cont(.75)  within group(order by percentagesales) over() as decimal(36,2)) as quartile75th
            ,cast(percentile_cont(1)  within group(order by percentagesales) over() as decimal(36,2)) as quartile100th
            from @sales
            )
            select 
                s.salesRepId
                ,s.percentageSales
                , case 
                    when s.percentageSales < q.quartile25th then 'red'
                    when s.percentageSales >=q.quartile25th and s.percentageSales <q.quartile75th then 'yellow'
                    when s.percentageSales >= q.quartile75th then 'green'
                 end as color

            From @sales s
            join quartile q on q.percentageSales = s.percentageSales and q.salesRepId = s.salesRepId

我如何像在 excel 中一样获得四分位数。

谢谢

Answer 2

在 percentageSales 中将 int 更改为 decimal：

declare @sales table(
    salesRepId int,
    percentageSales decimal(8,2))
    
    insert into @sales(salesRepId, percentageSales)
    values(1,40)
    ,(2,50)
    ,(3,50)
    ,(4,66.7)
    ,(5,83.33)
    ,(6,100)
    ,(7,100)
    ,(8,100)
    ,(9,100)
    ,(10,100)
    ,(11,100)
    ,(12,100)
    ,(13,100)
    ,(14,100)
    ,(15,100)
    ,(16,100)
    ,(17,100);
    
    with quintile as(
                select percentagesales, ntile(4) over(order by percentagesales) 
    as quintile
                from (select distinct percentagesales from @sales) as s
                )
                select salesrepid, r.percentagesales, q.quintile
                from @sales r
                join quintile q on r.percentagesales = q.percentagesales
                order by q.quintile, percentagesales