我有一个工作查询（oraclesql），它按存储编号对加仑数进行分组，并按类型和百分比列对加仑数进行求和。每个门店编号都有一个不同的转换日期，我从该日期开始汇总数据-

SELECT StoreNbr, 
SUM(CASE WHEN(ClrntSys IN ('844', '84448')) THEN Gallons ELSE 0 END) AS Gallons_844,
SUM(CASE WHEN(ClrntSys ='GIC') THEN Gallons ELSE 0 END) AS Gallons_GIC,
SUM(CASE WHEN(ClrntSys IN ('844', '84448', 'GIC')) THEN Gallons ELSE 0 END) AS Total_Gallons,
CONCAT(CAST((SUM(CASE WHEN(ClrntSys ='GIC') THEN Gallons ELSE 0 END) /
SUM(CASE WHEN(ClrntSys IN ('844', '84448', 'GIC')) THEN Gallons ELSE 0 END)) AS DECIMAL (5,2)) * 100, '%') AS Percent_GIC
FROM MQ_CDS_NETTRAN
WHERE ClrntSys IN ('844', '84448', 'GIC')
AND ((CostCenter = '701104' AND LastTranDate >= DATE '2020-03-10') 
OR (CostCenter = '701109' AND LastTranDate >= DATE '2020-03-04')
OR (CostCenter = '701257' AND LastTranDate >= DATE '2020-03-12'))
GROUP BY StoreNbr
ORDER BY StoreNbr;

我现在还需要每周总结一下，周日是每周的第一天，但我很难理解datepart是如何工作的。为了得到更好的想法，我只尝试用我在网上看到的datepart示例来总结上周的数据，但这不起作用。它给了我“无效的标识符日期部分”-

SELECT DATEPART(week, 5/17/20) AS weekTotal, 
StoreNbr, 
SUM(CASE WHEN(ClrntSys IN ('844', '84448')) THEN Gallons ELSE 0 END) AS Gallons_844,
SUM(CASE WHEN(ClrntSys ='GIC') THEN Gallons ELSE 0 END) AS Gallons_GIC,
SUM(CASE WHEN(ClrntSys IN ('844', '84448', 'GIC')) THEN Gallons ELSE 0 END) AS Total_Gallons,
CONCAT(CAST((SUM(CASE WHEN(ClrntSys ='GIC') THEN Gallons ELSE 0 END) /
SUM(CASE WHEN(ClrntSys IN ('844', '84448', 'GIC')) THEN Gallons ELSE 0 END)) AS DECIMAL (5,2)) * 100, '%') AS Percent_GIC
FROM MQ_CDS_NETTRAN
WHERE ((CostCenter = '701104' AND LastTranDate >= DATE '2020-03-10') 
OR (CostCenter = '701109' AND LastTranDate >= DATE '2020-03-04')
OR (CostCenter = '701257' AND LastTranDate >= DATE '2020-03-12'))
GROUP BY DATEPART(week, 5/17/20), StoreNbr
ORDER BY StoreNbr;

然而，我真正需要的是每周的数据（周日是每周的开始）分别汇总到每个商店的转换日期。有可能吗？除了datepart还有其他更好的方法吗？