显示在SQL Server中按聚合条件过滤的月份

Question

我有一个包含customer_id及其apply_date的表。我想在2016年的所有月份显示申请贷款的客户数量比当年的月平均客户数高30％。

customer_id   apply_date
-------------------------
1             2016-01-01
2             2016-02-01
3             2016-02-01
4             2016-02-01
5             2016-03-01
6             2016-03-01
7             2016-03-01
8             2016-03-01
9             2016-04-01
10            2016-05-01
11            2017-02-01
12            2017-02-01
13            2017-02-01

在此表中，2016年每月有2位客户申请（平均10位客户除以5个月）。比月平均值高30％的术语是乘以平均值2乘以1.3，我们得到2.6。

期望的结果是我想显示每个月有超过2.6个客户的月份。

通过使用2016年以上符合条件的表格仅为第2个月和第5个月。

上表只是数据的样本。

我尝试使用此代码

select 
    datepart(mm, apply_date) as month, count(*) as cnt
from 
    Leads
where 
    apply_date between '2016-01-01' and '2017-01-01'
group by 
    datepart(mm, apply_date)

但我不知道如何根据给定的条件过滤数据。

Answer 1

试试这段代码：

DECLARE @Leads TABLE
(
    customer_id    INT,
    apply_date DATETIME
)

INSERT INTO @Leads(customer_id, apply_date)VALUES(1,'2016-01-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(2,'2016-02-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(3,'2016-02-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(4,'2016-02-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(5,'2016-03-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(6,'2016-03-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(7,'2016-03-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(8,'2016-03-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(9,'2016-04-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(10,'2016-05-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(11,'2017-02-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(12,'2017-02-01');
INSERT INTO @Leads(customer_id, apply_date)VALUES(13,'2017-02-01');

--SELECT * FROM @Leads

DECLARE @avgInYear DECIMAL(18,2)

SELECT @avgInYear = SUM(a.NoOfCus)/ CAST(COUNT(a.MonthNo) AS DECIMAL(18,2)) FROM (
SELECT DISTINCT MonthNo = MONTH(apply_date), COUNT(customer_id) NoOfCus FROM @Leads
where apply_date between '2016-Jan-01' and '2017-Dec-01'
GROUP BY MONTH(apply_date) ) AS a

SELECT @avgInYear = @avgInYear * 1.3

select datepart(mm,apply_date) as month, COUNT(customer_id) as cnt
from @Leads
where apply_date between '2016-Jan-01' and '2017-Dec-01'
group by datepart(mm,apply_date)
HAVING CAST(COUNT(customer_id) AS DECIMAL(18,2)) > @avgInYear