最近工作中有需求，要從Excel表中提取一些統計資訊，因為需求變動可能比較頻繁，所以用VBA來做了。除錯資料操作部分頗費周折，特記錄一下。

1.資料來源檔案:[D:\4月份統計表.xlsx]：

2.最終的統計表格式：

3.計算規則：

要求：

A. 外觀等級A，且WLD在451.5-458之間，且LOP在82-200之間判定為OK，其他為NG。

B.將NG的按照表面等級→WLD→LOP的優先順序，判定是哪一項引起的不良，更新到資料來源的 [篩選結果] 中。

C.在統計表中按照入庫日期分組，統計總數量、OK數量、NG數量，以及NG產品中外觀等級、WLD、LOP各佔多大比例。

4.程式碼：

Option Explicit

Public Sub DataInquire()
    Dim strCnn As String, sqlStr As String, subSqlStr As String, strSql(1 To 4) As String '判斷條件 1=OK,2=外觀等級NG,3=WLDNG,4=LOPNG
    Dim MainBook As Workbook
    Dim Arr() As Variant, rngA As Range, rngB As Range
    Dim cnn As Object, rs As Object
    Dim i As Long, itemStr(1 To 4) As String '判定標誌 1=OK,2=外觀等級,3=WLD,4=LOP
    Dim sourceFile As String
    
    Set cnn = CreateObject("ADODB.Connection")
    Set rs = CreateObject("ADODB.Recordset")
    Set MainBook = ThisWorkbook
    sourceFile = "D:\4月份統計表.xlsx"
    strCnn = "Provider=Microsoft.ACE.OLEDB.12.0;Data Source='" + sourceFile + "';Extended Properties='Excel 12.0;HDR=YES,IME=1'"
        
    With ThisWorkbook.Worksheets(1)
        strSql(1) = " And [外觀等級] ='A' And [WLD] >=451.5 And [WLD] <=458 And LOP>=82 And LOP<200"
        strSql(2) = " And [外觀等級] <>'A'"
        strSql(3) = " And ([WLD] <451.5 or [WLD] >458)"
        strSql(4) = " And (LOP>=82 or LOP<200"
        itemStr(1) = "OK"
        itemStr(2) = "外觀等級"
        itemStr(3) = "WLD"
        itemStr(4) = "LOP"
        Set rngA = .Range("B" & .Cells(6000, "B").End(xlUp).Row + 1).Resize(1, 12) '//找到寫資料的位置
    End With
    
    cnn.Open strCnn
    If cnn.State = 1 Then
        For i = 4 To 1 Step -1 '//更新篩選結果
            sqlStr = "update [資料來源$] set [篩選結果]= '" & itemStr(i) & "' where 1=1 " & strSql(i)
            cnn.Execute sqlStr
        Next i
        '//////Excel/Access不支援case when then end結構，要用IIF來代替////
'        subSqlStr = "(select [序號], case [篩選結果] when 'OK' then 1 end as [OK]," & _
'                " case [篩選結果] when '外觀等級' then 1 end as [外觀等級]," & _
'                " case [篩選結果] when 'WLD' then 1 end as [WLD]," & _
'                " case [篩選結果] when 'LOP' then 1 end as [LOP]," & _
'                " from [資料來源$]) B "
'       ///////////////////////////////////////////////////////////////////
        subSqlStr = "(select [襯底編號], iif( [篩選結果] = 'OK',1,0 ) as [OK]," & _
                " iif( [篩選結果] = '外觀等級',1,0 ) as [外觀等級]," & _
                " iif( [篩選結果] = 'WLD',1,0 ) as [WLD]," & _
                " iif( [篩選結果] = 'LOP',1,0 ) as [LOP]," & _
                " from [資料來源$]) B "
        
        sqlStr = "select A.[入庫日期],count(A.[入庫日期]) as 總數量,sum(B.[OK]) as OK,(count(A.[入庫日期])-sum(B.[OK])) as NG,sum(B.[外觀等級])/(count(A.[入庫日期])-sum(B.[OK]))," & _
                " sum(B.[WLD])/(count(A.[入庫日期])-sum(B.[OK])),sum(B.[LOP])/(count(A.[入庫日期])-sum(B.[OK]))" & _
                "from [資料來源$] A, " & subSqlStr & " where A.[序號]=B.[序號] " & _
                "group by A.[入庫日期] order by A.[入庫日期] asc"
        Set rs = cnn.Execute(sqlStr)
        If Not rs.EOF Then
            Arr = WorksheetFunction.Transpose(rs.GetRows())
            Set rngA = rngA.Resize(UBound(Arr, 1), 12)
            rngA = Arr
            rngA.Borders.LineStyle = xlContinuous
            rngA.Resize(, 1).Offset(, 4).Resize(, 8).NumberFormatLocal = "0.00%"
        End If
        Set rs = Nothing
    End If
    cnn.Close
End Sub
 

A.要將[篩選結果]做細項分析，就要用到行轉列。而Excel/Access中沒有case when then end 結構，更沒有pivot可以用，幸好有IIF可用。

B.Transpose函式使用有些限制，要求結果集不超過1024行[未驗證]，且結果集不包含null[已驗證]。

C.Excel作為資料庫，查詢速度明顯比Access資料庫以及SqlServer、Oracle速度慢很多。5萬多行資料，要查詢大約20秒了。

D.Excel表作為資料庫使用，select和update可正常使用，但insert和delete不能用。使用select into 可以建立新的Excel工作表。