Question

这种情况似乎应该是一项简单的任务，但我无法提出解决方案。想象一下八列，其中四列是位列（开/关），另外四列是读数值。

我想从该表创建一个视图，其中另一列用于四个读取值的范围，基于其对应的传感器警报＆lt;＆gt;的那些值。 1 ..

出于示例目的，我们可以使用以下值。

第1栏 - 12.44
Column1Alarm - 0

第2栏 - 99.43
Column2Alarm - 0

第3栏 - 4.43
Column3Alarm - 1

第4栏 - 43.33
Column4Alarm - 0

对于此示例，范围中包含的唯一值应该是来自Column1,2和4的值。

谢谢，汤姆

Answer 1

根据@HLGM评论编辑，使其更加健壮。

请注意，在它的当前形式中，我假设

时

所有警报等于1，范围应为NULL
只有一个警报等于0，范围是此警报的值。

如果这还不够，OP可能会澄清应该返回的内容。

SQL语句

    ;WITH Alarm (C1, C1Alarm, C2, C2Alarm, C3, C3Alarm, C4, C4Alarm) AS (
        SELECT  12.44, 0, 99.43, 0, 4.43, 1, 43.33, 0
        UNION ALL SELECT 12.44, 1, 99.43, 0, 4.43, 1, 43.33, 0
        UNION ALL SELECT 1, 0, 2, 1, 3, 1, 4, 1
        UNION ALL SELECT 1, 1, 2, 1, 3, 1, 4, 1
    )
    , AddRowNumbers AS (
        SELECT  rowNumber = ROW_NUMBER() OVER (ORDER BY C1)
                , C1, C1Alarm
                , C2, C2Alarm
                , C3, C3Alarm
                , C4, C4Alarm
        FROM    Alarm   
    )
    , UnPivotColumns AS (
        SELECT  rowNumber, value = C1 FROM AddRowNumbers WHERE C1Alarm = 0
        UNION ALL SELECT rowNumber, C2 FROM AddRowNumbers WHERE C2Alarm = 0
        UNION ALL SELECT rowNumber, C3 FROM AddRowNumbers WHERE C3Alarm = 0
        UNION ALL SELECT rowNumber, C4 FROM AddRowNumbers WHERE C4Alarm = 0
    )
    SELECT  C1, C1Alarm
            , C2, C2Alarm
            , C3, C3Alarm
            , C4, C4Alarm
            , COALESCE(range1.range, range2.range)
    FROM    AddRowNumbers rowNumber
            LEFT OUTER JOIN (SELECT rowNumber, range = MAX(value) - MIN(value) FROM UnPivotColumns GROUP BY rowNumber HAVING COUNT(*) > 1) range1 ON range1.rowNumber = rowNumber.rowNumber
            LEFT OUTER JOIN (SELECT rowNumber, range = AVG(value) FROM UnPivotColumns GROUP BY rowNumber HAVING COUNT(*) = 1) range2 ON range2.rowNumber = rowNumber.rowNumber

测试脚本

;WITH Alarm (C1, C1Alarm, C2, C2Alarm, C3, C3Alarm, C4, C4Alarm) AS (
    SELECT  12.44, 0, 99.43, 0, 4.43, 1, 43.33, 0
    UNION ALL SELECT 12.44, 1, 99.43, 0, 4.43, 1, 43.33, 0
    UNION ALL SELECT 1, 0, 2, 1, 3, 1, 4, 1
    UNION ALL SELECT 1, 1, 2, 1, 3, 1, 4, 1
)
, AddRowNumbers AS (
    SELECT  rowNumber = ROW_NUMBER() OVER (ORDER BY C1)
            , C1, C1Alarm
            , C2, C2Alarm
            , C3, C3Alarm
            , C4, C4Alarm
    FROM    Alarm   
)
, UnPivotColumns AS (
    SELECT  rowNumber, value = C1 FROM AddRowNumbers WHERE C1Alarm = 0
    UNION ALL SELECT rowNumber, C2 FROM AddRowNumbers WHERE C2Alarm = 0
    UNION ALL SELECT rowNumber, C3 FROM AddRowNumbers WHERE C3Alarm = 0
    UNION ALL SELECT rowNumber, C4 FROM AddRowNumbers WHERE C4Alarm = 0
)
SELECT  C1, C1Alarm
        , C2, C2Alarm
        , C3, C3Alarm
        , C4, C4Alarm
        , COALESCE(range1.range, range2.range)
FROM    AddRowNumbers rowNumber
        LEFT OUTER JOIN (SELECT rowNumber, range = MAX(value) - MIN(value) FROM UnPivotColumns GROUP BY rowNumber HAVING COUNT(*) > 1) range1 ON range1.rowNumber = rowNumber.rowNumber
        LEFT OUTER JOIN (SELECT rowNumber, range = AVG(value) FROM UnPivotColumns GROUP BY rowNumber HAVING COUNT(*) = 1) range2 ON range2.rowNumber = rowNumber.rowNumber

Answer 2

OP的评论“即使我有200列......”让我相信在多个地方也需要类似的功能。所以我会创建一个接受“读取”列和“报警”位的函数，并在设置警报时返回NULL。这允许人们利用MIN和MAX处理NULL值的方式。

CREATE FUNCTION UnalarmedReading
(
    @Value float,
    @Alarm bit
)
RETURNS float
AS
BEGIN
return case when @Alarm=1 then null else @Value end
END

以下是一些测试数据：

create table Readings (
    KeyColumn int not null,
    Column1 float,
    Column1Alarm bit,
    Column2 float,
    Column2Alarm bit,
    Column3 float,
    Column3Alarm bit,
    Column4 float,
    Column4Alarm bit,
)

insert into Readings(
    KeyColumn,
    Column1,Column1Alarm,
    Column2,Column2Alarm,
    Column3,Column3Alarm,
    Column4,Column4Alarm
) values (
    1,
    12.44, 0,
    99.43, 0,
    4.43, 1,
    43.33, 0
)

insert into Readings(
    KeyColumn,
    Column1,Column1Alarm,
    Column2,Column2Alarm,
    Column3,Column3Alarm,
    Column4,Column4Alarm
) values (
    2,
    124.4, 0,
    994.3, 0,
    44.3, 1,
    433.3, 0
)

要使用MIN和MAX，请忽略：

;with NonAlarmReadings as (
    select KeyColumn, dbo.UnalarmedReading(Column1, Column1Alarm) as C1,
        dbo.UnalarmedReading(Column2, Column2Alarm) as C2,
        dbo.UnalarmedReading(Column3, Column3Alarm) as C3,
        dbo.UnalarmedReading(Column4, Column4Alarm) as C4
    from Readings
),
Normalized as (
    select *
    from NonAlarmReadings
    unpivot (Reading for BaseColumn in (C1, C2, C3, C4)) as upvt
)
select KeyColumn, min(Reading) as MinReading, max(Reading) as MaxReading,
    abs(min(Reading) - max(Reading)) as ReadingRange
from Normalized
group by KeyColumn

根据其他列值计算视图中的列

2 个答案:

SQL语句

测试脚本