Sql server SQL新手。我如何让下面的查询给出每日结果？

好的，我有下面的问题，但我正试着思考如何在给定的日期范围内给我每日结果。到目前为止，这基本上是我的疑问：

SELECT Source = CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                       WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                       WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                       WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                       ELSE Code END, Count(Document) as TotalCount, SUM(Amount) as TotalAmount                       

-- Joins here to gather and refine data collected

WHERE Date >= '03/20/2017'
       AND Date < '03/23/2017'
GROUP BY CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                           WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                           WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                           WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                           ELSE Code END
ORDER BY TotalCount DESC

理想情况下，它会给我：

Date         Source   TotalCount  TotalAmount
2017-03-20   SOURCE1  60          625
2017-03-20   SOURCE2  13          60
-- etc
2017-03-21   SOURCE1  63          600
2017-03-21   SOURCE2  10          66
-- etc

---

它还必须给出单个杂项来源（即除重新标记的主要4个来源外的1/+额外来源）的总计数和数量。我该怎么做呢？提前感谢您帮助一位新手

如果您只需将日期添加到select和group中，我想您应该会得到您想要的。像这样：

SELECT Date,Source = CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                       WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                       WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                       WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                       ELSE Code END, Count(Document) as TotalCount, SUM(Amount) as TotalAmount                       

-- Joins here to gather and refine data collected

WHERE Date >= '03/20/2017'
       AND Date < '03/23/2017'
GROUP BY CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                           WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                           WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                           WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                           ELSE Code END
ORDER BY date,TotalCount DESC

选择日期，当项目='code'或代码='1'时，选择Source=CASE，然后选择'SOURCE1'
当项目'CODE'和代码='2'时，则为'SOURCE2'
当项目'CODE'和代码='3'时，则为'SOURCE3'
当项目'CODE'和代码='4'时，则为'SOURCE4'
否则代码结束，将（文档）计数为TotalCount，将总和（金额）计数为TotalAmount
--在此联接以收集和优化收集的数据
其中日期>='2017年3月20日'
日期<'2017年3月23日'
当项目='CODE'或代码='1'然后是'SOURCE1'时按情况分组
当项目'CODE'和代码='2'时，则为'SOURCE2'
当项目'CODE'和代码='3'时，则为'SOURCE3'
当项目'CODE'和代码='4'时，则为'SOURCE4'
ELSE代码结束
按日期排序的订单，TotalCount DESC

听起来您希望在日期字段上进行聚合。因此，调整你的逻辑，使之成为这样，可能对你有用：

SELECT [Date]
, 'SOURCE' + Code AS Source
, COUNT(Document) as TotalCount
, SUM(Amount) as TotalAmount
--FROM...                       
GROUP BY [Date], 'SOURCE' + Code
ORDER BY COUNT(Document) DESC

这将为您提供与所请求内容更相似的输出。如果确实存在“Code”有时不等于1的实例，那么您可能需要添加一个额外的逻辑位。进行以下调整可能有效：

SELECT [Date]
, 'SOURCE' + CASE WHEN Project = 'CODE' OR Code = '1' THEN '1' ELSE Code END AS Source
, COUNT(Document) as TotalCount
, SUM(Amount) as TotalAmount     
--FROM...                  
GROUP BY [Date]
, 'SOURCE' + CASE WHEN Project = 'CODE' OR Code = '1' THEN '1' ELSE Code END
ORDER BY COUNT(Document) DESC

---

应该可以了。

@ChadPortman离我们很近，只是在分组中遗漏了日期。最终查询如下所示：

SELECT Date, Source = CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                   WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                   WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                   WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                   ELSE Code END, Count(Document) as TotalCount, SUM(Amount) as TotalAmount                       

-- Joins here to gather and refine data collected

WHERE Date >= '03/20/2017'
       AND Date < '03/23/2017'
GROUP BY Date, CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                           WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                           WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                           WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                           ELSE Code END
ORDER BY Date, TotalCount DESC

选择日期，当项目='code'或代码='1'时，选择Source=CASE，然后选择'SOURCE1'
当项目'CODE'和代码='2'时，则为'SOURCE2'
当项目'CODE'和代码='3'时，则为'SOURCE3'
当项目'CODE'和代码='4'时，则为'SOURCE4'
否则代码结束，将（文档）计数为TotalCount，将总和（金额）计数为TotalAmount
--在此联接以收集和优化收集的数据
其中日期>='2017年3月20日'
日期<'2017年3月23日'
按日期分组，当项目='CODE'或代码='1'然后是'SOURCE1'时按大小写
当项目'CODE'和代码='2'时，则为'SOURCE2'
当项目'CODE'和代码='3'时，则为'SOURCE3'
当项目'CODE'和代码='4'时，则为'SOURCE4'
ELSE代码结束
按日期排序的订单，TotalCount DESC

---

这正是我所需要的。谢谢大家的回复

Sql不会动态生成列。在开始编写查询之前，您需要知道每个列名和类型的确切值。谢谢您的提醒，@JoelCoehoorn。正在更新所需的输出，使其更好地工作。您还需要在日期上执行分组操作，以获得所需的结果完全从我身边飞过，按日期对其进行分组。。。我需要更多的咖啡。明白了，@Rams！我正要回复我的帖子，回答是这样的。谢谢谢谢你的回复！

SELECT Date, Source = CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                   WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                   WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                   WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                   ELSE Code END, Count(Document) as TotalCount, SUM(Amount) as TotalAmount                       

-- Joins here to gather and refine data collected

WHERE Date >= '03/20/2017'
       AND Date < '03/23/2017'
GROUP BY Date, CASE WHEN Project = 'CODE' OR Code = '1' THEN 'SOURCE1'
                           WHEN Project <> 'CODE' AND Code = '2' THEN 'SOURCE2'
                           WHEN Project <> 'CODE' AND Code = '3' THEN 'SOURCE3'
                           WHEN Project <> 'CODE' AND Code = '4' THEN 'SOURCE4'
                           ELSE Code END
ORDER BY Date, TotalCount DESC