SQL - zwracane skumulowane miesięczne wartości kolumn

Question

Mam to zapytanie SQL, którego używam do znajdowania ruchów próbnych na konto miesięcznie.

/* Fixed Assets 1 */ 
SELECT * FROM 
(
    SELECT T0.AcctCode AS 'SAP Code', T0.AcctName AS 'Description', 
    MONTH(T1.RefDate) AS Month, SUM(T1.Debit - T1.Credit) AS 'Amount' 
    FROM OACT T0 
    LEFT JOIN JDT1 T1 ON T0.[AcctCode] = T1.[Account] 
    WHERE T0.AcctCode LIKE '111%' AND T0.Levels = 5 
    AND (T1.RefDate BETWEEN DATEADD(yy, DATEDIFF(yy,0,{?AsAtDate}), 0) AND {?AsAtDate}) 
    GROUP BY T0.AcctCode, T0.AcctName, T0.FatherNum, T1.RefDate) AS q 
    PIVOT 
(
SUM(Amount) 
FOR [Month] IN ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12]) 
) AS query 

ta zwraca następujące wyniki od czerwca:

Chciałbym obejmują łączne miesięczne wyniki dla wszystkich rzędach tak, że rząd 1 wyniki powinny być:

1, 111110, Building Gross Value, 633604.23, 637764.23, 645313.03, 649061.78, 651097.78, 651097.78, 651397.78 

Nie udało mi się tego osiągnąć. Jak zmodyfikować zapytanie, aby to osiągnąć?

Dodatkowe dane:

zapytaniu:

SELECT CAST(T0.TaxDate AS Date), T0.Account, T1.AcctName, T0.Debit, T0.Credit 
FROM JDT1 T0 INNER JOIN OACT T1 
ON T0.Account = T1.AcctCode 
WHERE T0.[Account] = '111110' 

Powroty:

Answer 1

Trzeba zewnętrzna select w których dodać kolumny do drugiego:

SELECT AcctCode AS 'SAP Code', AcctName AS 'Description', 
col1 AS '1', 
col1 + col2 AS '2', 
col1 + col2 + col3 AS '3', 
... FROM (
    SELECT AcctCode, AcctName, 
    [1] AS col1, 
    [2] AS col2, 
    [3] AS col3, 
    ... FROM 
    (
     SELECT T0.AcctCode, T0.AcctName, 
     MONTH(T1.RefDate) AS Month, SUM(T1.Debit - T1.Credit) AS 'Amount' 
     FROM OACT T0 
     LEFT JOIN JDT1 T1 ON T0.[AcctCode] = T1.[Account] 
     WHERE T0.AcctCode LIKE '111%' AND T0.Levels = 5 
     AND (T1.RefDate BETWEEN DATEADD(yy, DATEDIFF(yy,0,{?AsAtDate}), 0) AND {?AsAtDate}) 
     GROUP BY T0.AcctCode, T0.AcctName, T0.FatherNum, T1.RefDate) AS q 
     PIVOT 
    (
    SUM(Amount) 
    FOR [Month] IN ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12]) 
    ) AS query 
) 

Answer 2

Najbardziej bezpośrednim sposobem jest zrzucić wyniki przegubu do tabeli tymczasowej lub, jeśli zbiór danych jest dość mały, zmienna tabeli. Następnie użyj UNION, aby uzyskać skumulowane wyniki i podkategorie z PIVOT. Oto przykład użycia zmiennej tabeli.

/* Fixed Assets 1 */ 
SELECT * 
INTO #FixedAssets FROM 
(
    SELECT T0.AcctCode AS 'SAP Code', T0.AcctName AS 'Description', 
    MONTH(T1.RefDate) AS Month, SUM(T1.Debit - T1.Credit) AS 'Amount' 
    FROM OACT T0 
    LEFT JOIN JDT1 T1 ON T0.[AcctCode] = T1.[Account] 
    WHERE T0.AcctCode LIKE '111%' AND T0.Levels = 5 
    AND (T1.RefDate BETWEEN DATEADD(yy, DATEDIFF(yy,0,{?AsAtDate}), 0) AND {?AsAtDate}) 
    GROUP BY T0.AcctCode, T0.AcctName, T0.FatherNum, T1.RefDate) AS q 
    PIVOT 
(
SUM(Amount) 
FOR [Month] IN ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12]) 
) AS query 
GO 

SELECT 
     [SAP Code] = '111110' 
    , [Description] = 'Building Gross Value' 
    , [1]   = SUM([1]) 
    , [2]   = SUM([2]) 
    , [3]   = SUM([3]) 
    , [4]   = SUM([4]) 
    , [5]   = SUM([5]) 
    , [6]   = SUM([6]) 
    , [7]   = SUM([7]) 
    , [8]   = SUM([8]) 
    , [9]   = SUM([9]) 
    , [10]   = SUM([10]) 
    , [11]   = SUM([11]) 
    , [12]   = SUM([12]) 
UNION 
SELECT * FROM #FixedAssets 
ORDER BY 1 

Answer 3

Gdybym zrozumiał swoje pytanie poprawnie, trzeba skumulowane wyniki miesięczne w jednym rzędzie z przecinkiem oddzielone Obciążenia wartości

create table a 
(dateval date, acc int,accname varchar(100), debit decimal(10,2)) 

insert into a 
values 
('20150630', 111110, 'Building Gross Value', 633604.230), 
('20150731', 111110, 'Building Gross Value', 2760.000000), 
('20150730', 111110, 'Building Gross Value', 1400.000000), 
('20150808', 111110, 'BUIIdlngGrossValue', 1890.00), 
('20150811', 111110, 'BUIIdlnanossValue', 180.00), 
('20150811', 111110, 'Building Gross Value', 375.000000), 
('20150819', 111110, 'Building Gross Value', 2622.200000), 
('20150821', 111110, 'Building Gross Value', 360.000000), 
('20150822', 111110, 'Building Gross Value', 21.600000), 
('20150824', 111110, 'Building Gross Value', 100.000000), 
('20150825', 111110, 'Building Gross Value', 770.000000), 
('20150829', 111110, 'Building Gross Value', 100.000000), 
('20150831', 111110, 'Building Gross Value', 340.000000), 
('20150831', 111110, 'Building Gross Value', 790.000000), 
('20150924', 111110, 'Building Gross Value' ,918.750000), 
('20150928', 111110, 'Building Gross Value', 2830.000000), 
('20151005', 111110, 'Building Gross Value', 1411.000000), 
('20151023', 111110, 'Building Gross Value', 625.000000) 

Teraz poniżej wymienione zapytanie daje wyniki w jednym wierszu z wartościami oddzielonymi przecinkami.

with cte as 
(
select * from 
(
    select a.dateval,b.acc,b.accname,sum(b.debit) as debit,row_number() over (partition by year(a.dateval),month(a.dateval) order by a.dateval desc) as rn 
    from (select distinct dateval from a) as a 
    inner join a as b 
    on a.dateval >= b.dateval 
    group by a.dateval,b.acc,b.accname 
) as a 
where rn = 1 
) 
select acc,accname,stuff((select ',' + cast(debit as varchar(1000)) from cte as b where a.acc = b.acc for xml path ('')),1,1,'') 
from cte as a 
group by acc,accname