TSQL資料透視表行到列

我在一個表中得到了一些資料，如下所示。我正在嘗試運行一個查詢，該查詢將在每個 requestId 的一行中獲取我的資料。我不需要日期或拒絕原因，只需要每個 groupId 的 eprojman 和 apvStatus

requestId - projMan1 - apvStatus1 - projMan2 - apvStatus2 - projMan3 - apvStatus3 等等。對于所有 5 個 groupIds

我一直在嘗試使用 PIVOT 表，但我發現的所有示例都涉及匯總資料或其他內容。我只是很想取 5 行并將其變成 1 為每個 requestID

我唯一能想到的是從同一個表中為每個 groupID 選擇 5 次并將其合并，但這比見鬼的要慢。必須有更好的方式

謝謝。

當前查詢：

select group1.requestId
    , group1.apvStatus as apvStatus1
    , group1.projMan as projMan1
    , group2.apvStatus as apvStatus2
    , group2.projMan as projMan2
    , group3.apvStatus as apvStatus3
    , group3.projMan as projMan3
    ,group4.apvStatus as apvStatus4
    , group4.projMan as projMan4
    ,group5.apvStatus as apvStatus5
    , group5.projMan as projMan5
    ,group1.denialReason 
    INTO #TEMPBAOrganized
    from (
    select requestId, apvStatus, projMan, denialReason from #TEMPBULKAPPROVAL where groupId = 1) group1
    INNER JOIN 
    (select requestId, apvStatus, projMan, denialReason from #TEMPBULKAPPROVAL where groupId = 2) group2
    on group1.requestId = group2.requestId
    INNER JOIN
    (select requestId, apvStatus, projMan from #TEMPBULKAPPROVAL where groupId = 3) group3
    on group1.requestId = group3.requestId
    INNER JOIN
    (select requestId, apvStatus, projMan from #TEMPBULKAPPROVAL where groupId = 4) group4
    on group1.requestId = group4.requestId
    INNER JOIN
    (select requestId, apvStatus, projMan from #TEMPBULKAPPROVAL where groupId = 5) group5
    on group1.requestId = group5.requestId

uj5u.com熱心網友回復：

對于多列的旋轉，使用CASE帶有聚合的運算式更容易。

select  t.requestId,
        projMan1    = max(case when t.groupId = 1 then t.projMan end),
        apvStatus1  = max(case when t.groupId = 1 then t.apvStatus end),
        projMan2    = max(case when t.groupId = 2 then t.projMan end),
        apvStatus2  = max(case when t.groupId = 2 then t.apvStatus end),
        projMan3    = max(case when t.groupId = 3 then t.projMan end),
        apvStatus3  = max(case when t.groupId = 3 then t.apvStatus end),
        projMan4    = max(case when t.groupId = 4 then t.projMan end),
        apvStatus4  = max(case when t.groupId = 4 then t.apvStatus end),
        projMan5    = max(case when t.groupId = 5 then t.projMan end),
        apvStatus5  = max(case when t.groupId = 5 then t.apvStatus end)
from    #TEMPBULKAPPROVAL t
group by t.requestId

注意：max也是一個聚合函式

uj5u.com熱心網友回復：

你可以這樣做：

SELECT
    requestId,
    approvalDate_1 = MAX(approvalDate_1),
    approvalDate_2 = MAX(approvalDate_2),
    approvalDate_3 = MAX(approvalDate_3),
    approvalDate_4 = MAX(approvalDate_4),
    approvalDate_5 = MAX(approvalDate_5), 
    projMan_1 = MAX(projMan_1),
    projMan_2 = MAX(projMan_2),
    projMan_3 = MAX(projMan_3),
    projMan_4 = MAX(projMan_4),
    projMan_5 = MAX(projMan_5)
FROM
(
SELECT
    requestId,
    groupId, 
    approvalDate,
    projMan,
    'approvalDate_'   CAST( groupId AS VARCHAR(2)) AS approvalDatePivot,
    'projMan_'   CAST( groupId AS VARCHAR(2)) AS projManPivot
FROM
  @tbl
) T
PIVOT (
  MAX(approvalDate) FOR approvalDatePivot IN ([approvalDate_1],[approvalDate_2],[approvalDate_3],[approvalDate_4],[approvalDate_5])
) pvt_1
PIVOT (
  MAX(projMan) FOR projManPivot IN ([projMan_1],[projMan_2],[projMan_3],[projMan_4],[projMan_5])
) pvt_2

GROUP BY requestId

標籤：sql sql-server 查询语句

上一篇：如何提高我的查詢性能SQLServer

下一篇：具有相同多個子查詢的SQL陳述句運行速度很慢