在这种情况下，如何 select nvarchar 数据类型的不同值并给它们排名？

Question

我是运行 SQL Server 2012，我的数据库中有一个名为 DateDimension 的 table。我的 DateDimension Table 是一个包含 2 列的 table：第 date 列包含从“2014-01-01”到“[=20=”的每日日期运行 ]' 和列 WeekGroup 是 nvarchar(100) 数据类型。

摘录如下：

date         WeekGroup
2014-01-01   Wk 30 Dec 2013-05 Jan 2014
2014-01-02   Wk 30 Dec 2013-05 Jan 2014
2014-01-03   Wk 30 Dec 2013-05 Jan 2014
2014-01-04   Wk 30 Dec 2013-05 Jan 2014
2014-01-05   Wk 30 Dec 2013-05 Jan 2014
2014-01-06   Wk 06 Jan 2014-12 Jan 2014
...

我现在有以下查询：

SELECT
[date],
[WeekGroup],
RANK() OVER (ORDER BY [date] ASC) AS 'Rank'

FROM DateDimension

WHERE date BETWEEN '2014-01-01' AND '2019-12-31'

ORDER BY [date] ASC

我的输出如下：

date         WeekGroup                     Rank
2014-01-01   Wk 30 Dec 2013-05 Jan 2014     1
2014-01-02   Wk 30 Dec 2013-05 Jan 2014     2
2014-01-03   Wk 30 Dec 2013-05 Jan 2014     3
2014-01-04   Wk 30 Dec 2013-05 Jan 2014     4
2014-01-05   Wk 30 Dec 2013-05 Jan 2014     5
2014-01-06   Wk 06 Jan 2014-12 Jan 2014     6
...

我想微调上述查询，以便获得以下输出，其中 distinct WeekGroup 和相应的 Rank:

WeekGroup                     Rank
Wk 30 Dec 2013-05 Jan 2014     1
Wk 06 Jan 2014-12 Jan 2014     2
...

如何在 nvarchar 变量 (WeekGroup) 上执行此操作？

Answer 1

将 date 排除在外并按周分组汇总：

SELECT WeekGroup,
       RANK() OVER (ORDER BY MIN([date]) ASC) AS [Rank]
FROM DateDimension
WHERE date BETWEEN '2014-01-01' AND '2019-12-31'
GROUP BY WeekGroup
ORDER BY MIN([date]) ASC;

Answer 2

这使用了 CTE 和 RANK()：

declare @DateDimension table ([date] date, WeekGroup nvarchar(100))
insert into @DateDimension values
  ('2014-01-01', 'Wk 30 Dec 2013-05 Jan 2014')
, ('2014-01-02', 'Wk 30 Dec 2013-05 Jan 2014')
, ('2014-01-03', 'Wk 30 Dec 2013-05 Jan 2014')
, ('2014-01-04', 'Wk 30 Dec 2013-05 Jan 2014')
, ('2014-01-05', 'Wk 30 Dec 2013-05 Jan 2014')
, ('2014-01-06', 'Wk 06 Jan 2014-12 Jan 2014')

;with cte as (
SELECT [WeekGroup], count(*) as count
FROM @DateDimension
WHERE date BETWEEN '2014-01-01' AND '2019-12-31'
group by [WeekGroup])
select [WeekGroup], rank() over (order by [count] desc) as [Rank] from cte

Answer 3

试试这个。

   declare @DatA table ([date] date, WeekGroup nvarchar(100))
    insert into @DatA values
      ('2014-01-01', 'Wk 30 Dec 2013-05 Jan 2014')
    , ('2014-01-02', 'Wk 30 Dec 2013-05 Jan 2014')
    , ('2014-01-03', 'Wk 30 Dec 2013-05 Jan 2014')
    , ('2014-01-04', 'Wk 30 Dec 2013-05 Jan 2014')
    , ('2014-01-05', 'Wk 30 Dec 2013-05 Jan 2014')
    , ('2014-01-06', 'Wk 06 Jan 2014-12 Jan 2014')
    SELECT WEEKGROUP, RANK() OVER (ORDER BY A.WeekGroup desc) RNK
    FROM (
              SELECT DISTINCT WEEKGROUP FROM @DatA
              WHERE date BETWEEN '2014-01-01' AND '2019-12-31'
         )A

在这种情况下，如何 select nvarchar 数据类型的不同值并给它们排名？

How can I select Distinct Values of nvarchar datatype and give them a Rank in this scenario?

sql

sql-server

nvarchar

distinct

rank