anzeigen SQL Ergebnis im Querformat

Question

Alles, was ich will, ist diese zu transformieren:

Period | Department | Print | Copy 
--------------------------------------- 
201601 | Dept 1  | 10 | 20 
201601 | Dept 2  | 20 | 10 
201602 | Dept 1  | 30 | 40 
201602 | Dept 2  | 40 | 30 
201603 | Dept 1  | 50 | 60 
201603 | Dept 2  | 60 | 50 

in dieser:

Department | 201601 Print | 201601 Copy | 201602 Print | 201602 Copy | 201603 Print | 201603 Copy 
------------------------------------------------------------------------------------------ 
Dept 1  | 10   | 20   | 30   | 40   | 50   | 60 
Dept 2  | 20   | 10   | 40   | 30   | 60   | 50 

Ich habe versucht, das Skript mit PIVOT zu bauen, aber ich weiß nicht, wie um sowohl "Drucken" als auch "Kopieren" für jede Periode in den Spalten anzuzeigen. Außerdem, da die Werte von 'Periode' unbekannt sind, kann ich den Wert im Skript nicht fest codieren.

Hier ist mein Versuch:

SELECT [Department] 
    ,[201601] AS [201601 Copy] 
    ,[201602] AS [201602 Copy] 
    ,[201603] AS [201603 Copy] 
FROM 
    (SELECT [Copy], [Period], [Department] from #tempTable) AS ST 
PIVOT 
    (SUM([Copy]) FOR [Period] IN ([201601],[201602],[201603])) AS PT 

Und hier ist das Skript für die Tabelle mit meiner Beispieldaten zu erstellen:

IF OBJECT_ID('tempdb..#tempTable') IS NOT NULL 
    DROP TABLE #tempTable 
CREATE TABLE #tempTable(
    [Period] varchar(50) 
    ,[Department] varchar(50) 
    ,[Print] int 
    ,[Copy] int 
) 
INSERT INTO #tempTable VALUES 
    ('201601', 'Dept 1', 10, 20) 
    ,('201601', 'Dept 2', 20, 10) 
    ,('201602', 'Dept 1', 30, 40) 
    ,('201602', 'Dept 2', 40, 30) 
    ,('201603', 'Dept 1', 50, 60) 
    ,('201603', 'Dept 2', 60, 50) 

Vielen Dank für jede Antwort im Voraus.

ANTWORT

Ich studierte die eingegangenen Antworten und schließlich die unten stehende Skript bauen:

DECLARE @sql AS varchar(max); 
SELECT @sql = 'SELECT [Department],' + 
    STUFF((
     SELECT DISTINCT 
      ',SUM(ISNULL(CASE [Period] WHEN ''' + [Period] + ''' THEN [Print] END, 0)) AS [' + [period] + ' Print]' + 
      ',SUM(ISNULL(CASE [Period] WHEN ''' + [Period] + ''' THEN [Copy] END, 0)) AS [' + [period] + ' Copy]' 
     FROM #TempTable 
     FOR XML PATH('') 
    ), 1, 1, '') + 
    'FROM #TempTable 
    GROUP BY [Department]'; 
PRINT @sql 
EXEC(@sql); 

Answer 1

Sie eine dynamische SQL-Abfrage verwenden kann.

Abfrage

declare @sql as varchar(max); 
select @sql = 'select [Department],' + stuff((
    select distinct ',max(case [Period] when ' + char(39) + [Period] + char(39) + 
    ' then [Print] end) [' + [period] + ' Print]' 
    + ',max(case [Period] when ' + char(39) + [Period] + char(39) + 
    ' then [Copy] end) [' + [period] + ' Copy]' 
    from #TempTable 
    for xml path('') 
), 1, 1, ''); 

select @sql += ' from #TempTable group by [Department];'; 
exec(@sql); 

Answer 2

Sie können diese mit ISNULL() und SUM() Funktionen erreichen.

SELECT [Department] 
     ,SUM(ISNULL(CASE WHEN [Period]='201601' THEN [Print] END,0)) AS [201601 Print] 
     ,SUM(ISNULL(CASE WHEN [Period]='201601' THEN Copy END,0)) AS [201601 Copy] 
     ,SUM(ISNULL(CASE WHEN [Period]='201602' THEN [Print] END,0)) AS [201602 Print] 
     ,SUM(ISNULL(CASE WHEN [Period]='201602' THEN Copy END,0)) AS [201602 Copy] 
     ,SUM(ISNULL(CASE WHEN [Period]='201603' THEN [Print] END,0)) AS [201603 Print] 
     ,SUM(ISNULL(CASE WHEN [Period]='201603' THEN Copy END,0)) AS [201603 Copy]    
FROM #tempTable 
GROUP BY [Department] 

Answer 3

SELECT Department,SUM([201601Print])[201601 Print],SUM([201601Copy])[201601 Copy],SUM([201602Print])[201602 Print], 
       SUM([201602Copy])[201602 Copy],SUM([201603Print])[201603 Print],SUM([201603Copy])[201603 Copy] FROM (
SELECT [Department] 
,[201601] AS [201601Copy] 
,[201602] AS [201602Copy] 
,[201603] AS [201603Copy] 
,0 AS [201601Print] 
,0 AS [201602Print] 
,0 AS [201603Print] 
FROM 
(SELECT [Period],[Copy], [Department] from #tempTable) AS ST 
PIVOT 
(SUM([Copy]) FOR [Period] IN ([201601],[201602],[201603])) AS PT 
UNION ALL 
SELECT [Department] 
,0 AS [201601Copy] 
,0 AS [201602Copy] 
,0 AS [201603Copy] 
,[201601] AS [201601Print] 
,[201602] AS [201602Print] 
,[201603] AS [201603Print] 
FROM 
(SELECT [Period],[Print], [Department] from #tempTable) AS ST 
PIVOT 
(SUM([Print]) FOR [Period] IN ([201601],[201602],[201603])) AS PT 
)A GROUP BY Department 

Answer 4

Eine weitere dynamische SQL, die einen Drehpunkt verwendet.
Aber dieser verwendet eine Variable @ Columns, um eine Zeichenfolge mit den Spaltennamen zu generieren.

declare @Columns varchar(max); 
set @Columns = STUFF((SELECT ', ' + QUOTENAME([Period] +' Print') + ', ' + QUOTENAME([Period] +' Copy') FROM #tempTable GROUP BY [Period] FOR XML PATH(''), TYPE).value('.', 'NVARCHAR(max)') ,1,1,''); 

declare @SQL varchar(max); 
set @SQL = 'select * 
from (
select [Department], [Period] + '' Print'' as Title, [Print] as Value from #tempTable 
union all 
select [Department], [Period] + '' Copy'' as Title, [Copy] as Value from #tempTable 
) q 
pivot (sum(Value) for Title in ('+ @Columns +')) p;'; 

--select @SQL; 
exec (@SQL);