查询循环数据并为每个唯一值提供 Table

Question

我正在尝试编写一个查询来循环遍历来自每种类型的特定 table 和 return 结果的数据。我已经弄清楚了一些逻辑我想无论如何我只知道它不会工作因为我遗漏了一些东西。

我正在使用的数据结构如下：

        Customer ID     Delivery Type
    143             Delayed
    123             Delivered 
    14              In Process
    123             In Route
    432             Delivered 
    456             Delayed
    76              In Route
    34              In Route
    546             Delayed
    324             Delivered 
    235             In Process
    678             Delayed
    234             In Route
    123             In Route
    321             Delayed
    987             In Process
    546             In Process
    324             Delayed
    235             In Route
    678             In Process
    234             Delayed
    123             In Route
    321             Delivered 
    987             In Process

我试图得到这样的结果：

Delivery Type   Customer ID  Count
Delayed         143             1
Delayed         234             1
Delayed         321             1
Delayed         324             1
Delayed         456             1
Delayed         546             1
Delayed         678             1

Delivery Type   Customer ID  Count
Delivered       123             1
Delivered       321             1
Delivered       324             1
Delivered       432             1

Delivery Type   Customer ID  Count
In Process      14              1
In Process      235             1
In Process      546             1
In Process      678             1
In Process      987             2

Delivery Type   Customer ID  Count
In Route        34              1
In Route        76              1
In Route        123             3
In Route        234             1
In Route        235             1

我从以下我知道行不通的逻辑开始：

DECLARE @i int
DECLARE @DeliveryCount int
DECLARE @Type int

@type = Select distinct DeliveryType from Deliveries

@DeliveryCount = select count(*) as 'Count' from @type

SET @i = 0

WHILE @DeliveryCount > @i
BEGIN 
    @DeliveryType = Select row @i from @type

    @Top10 = SELECT TOP 10 count(*) as 'Count', CustomerID, DeliveryType FROM Deliveries
            Group by CustomerID, DeliveryType
            Where DeliveryType like @DeliveryType
    Print @Top10

    Set @i = @i + 1
END

Answer 1

你似乎想要 group by 和 order by:

select delivery_type, customer_id, count(*)
from deliveries
group by delivery_type, customer_id
order by delivery_type, customer_id;

我不知道您为什么要尝试循环执行此操作。您应该只在 SQL.

中使用基于集合的逻辑

编辑：

根据您的评论，使用 window 函数：

select d.*
from (select delivery_type, customer_id,
             count(*) as cnt,
             row_number() over (partition by delivery_type order by count(*) desc) as seqnum
      from deliveries
      group by delivery_type, customer_id
     ) d
where seqnum <= 10
order by delivery_type, customer_id;

Answer 2

您可以使用聚合和 window 函数按交付类型获取前 10 位客户：

select *
from (
    select 
        DeliveryType, 
        CustomerID, 
        count(*),
        rank() over(partition by DeliveryType order by count(*) desc) rn
    from Deliveries
    group by DeliveryType, CustomerID
) t
where rn <= 10
order by DeliveryType, CustomerID, rn;

或者，您也可以使用 top 1 with ties - 这消除了对子查询的需要，但不允许您对结果进行排序：

select top (1) with ties
    DeliveryType, 
    CustomerID, 
    count(*)
from Deliveries
group by DeliveryType, CustomerID
order by (rank() over(partition by DeliveryType order by count(*) desc) - 1) / 10

查询循环数据并为每个唯一值提供 Table

Query to Loop Through Data and Provide Table for Each Unique Value

sql

tsql

sql-server

group-by

greatest-n-per-group