sql-server - 多个枢轴？需要按每月小时数计算

Question

这应该很简单。我有其他 PIVOT SQL 查询工作正常。我想统计登录次数:按小时、按月。我正在考虑两个 PIVOT 或 UNPIVOT 然后再 PIVOT？是的，我已经在这里、其他网站、Google 等进行了挖掘。我陷入了困境。

SELECT
loginid
,DATEPART(MONTH,logtime) Month
, DATEPART(HOUR, logtime) Hour
FROM somelog (nolock)
) temp
PIVOT (
COUNT(loginid)
        FOR Month in (JAN,FEB,MAR,APR,MAY,JUN,JUL,AUG,SEP,OCT,NOV,DEC)
) AS Pvt

我想要的结果是什么..

一月、二月、三月小时

00

01

02

..

23

而且我不需要 8760 (365 x 24) 个

我尝试过按小时分组 我尝试过 GROUP BY temp.hour

我也尝试过这个。它似乎确实有效，但我得到了 00 小时，例如 365 次。再次，GROUP BY 问题..

SELECT
TimeOfDay
, [1] JAN
, [2] FEB
, [3] MAR
, [4] APR
, [5] MAY
, [6] JUN
, [7] JUL
, [8] AUG
, [9] SEP
, [10] OCT
, [11] NOV
, [12] DEC
FROM (SELECT logintime
, loginid
, datepart(month, logintime) [month]
, DatePart(hour, logintime) TimeOfDay
FROM sometable (nolock)) x
PIVOT (
COUNT(loginid)
for [month] in ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12])
) AS pvt
ORDER BY 1, 2

谢谢

肯特

Answer 1

一种方法

WITH hour_list AS (
  SELECT 0 hour
  UNION ALL
  SELECT hour + 1 FROM hour_list WHERE hour < 23
)
SELECT h.hour,
       COALESCE(jan, 0) jan,
       COALESCE(feb, 0) feb,
       COALESCE(mar, 0) mar,
       COALESCE(apr, 0) apr,
       COALESCE(may, 0) may,
       COALESCE(jun, 0) jun,
       COALESCE(jul, 0) jul,
       COALESCE(aug, 0) aug,
       COALESCE(sep, 0) sep,
       COALESCE(oct, 0) oct,
       COALESCE(nov, 0) nov,
       COALESCE(dec, 0) dec
  FROM hour_list h LEFT JOIN
(
  SELECT DATEPART(HOUR, logtime) hour, 
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  1 THEN 1 END) jan,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  2 THEN 1 END) feb,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  3 THEN 1 END) mar,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  4 THEN 1 END) apr,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  5 THEN 1 END) may,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  6 THEN 1 END) jun,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  7 THEN 1 END) jul,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  8 THEN 1 END) aug,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  9 THEN 1 END) sep,
         SUM(CASE WHEN DATEPART(MONTH, logtime) = 10 THEN 1 END) oct,
         SUM(CASE WHEN DATEPART(MONTH, logtime) = 11 THEN 1 END) nov,
         SUM(CASE WHEN DATEPART(MONTH, logtime) = 12 THEN 1 END) dec
    FROM somelog (NOLOCK)
   GROUP BY DATEPART(HOUR, logtime)
) l 
    ON h.hour = l.hour

或使用PIVOT

WITH hour_list AS (
  SELECT 0 hour
  UNION ALL
  SELECT hour + 1 FROM hour_list WHERE hour < 23
)
SELECT h.hour,
       COALESCE([1],  0) jan,
       COALESCE([2],  0) feb,
       COALESCE([3],  0) mar,
       COALESCE([4],  0) apr,
       COALESCE([5],  0) may,
       COALESCE([6],  0) jun,
       COALESCE([7],  0) jul,
       COALESCE([8],  0) aug,
       COALESCE([9],  0) sep,
       COALESCE([10], 0) oct,
       COALESCE([11], 0) nov,
       COALESCE([12], 0) dec
  FROM hour_list h LEFT JOIN 
(
  SELECT DATEPART(MONTH, logtime) month,
         DATEPART(HOUR, logtime) hour,
         COUNT(*) log_count
    FROM somelog (NOLOCK)
   GROUP BY DATEPART(MONTH, logtime), DATEPART(HOUR, logtime)
) s
PIVOT
(
  SUM(log_count) FOR month IN([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12])
) p
    ON h.hour = p.hour

两个查询的示例输出:

| HOUR | JAN | FEB | MAR | APR | MAY | JUN | JUL | AUG | SEP | OCT | NOV | DEC |
|------|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
|    0 |   2 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    1 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    2 |   0 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    3 |   0 |   0 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    4 |   0 |   0 |   0 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
...

这里是SQLFiddle 演示(使用 CASE)
这是SQLFiddle 演示(使用 PIVOT)