MySQL - 计数呼叫在特定范围内 0 到 10 秒内应答，按计数分组答案

【问题标题】：MySQL - Count call answered within certain ranges 0 to 10 sec, group by countMySQL - 计数呼叫在特定范围内 0 到 10 秒内应答，按计数分组
【发布时间】：2017-09-11 08:12:46
【问题描述】：

我需要计算在 0 到 10 秒、0 到 20 秒等特定范围内应答的呼叫。计数会增加，而增量将是当前和下一个之间的不同计数。 % 将是当前计数/最终计数总数。

这是您可以使用数据进行测试的 sqlfiddle：http://sqlfiddle.com/#!9/803d2/2

调用详情示例表：

+-----+----------------+----------+----------+---------------+
| id  | callid         | callerno | duration | status        |
+-----+----------------+----------+----------+---------------+
| 634 | 1479097551.228 | 1000     |        2 | complete      |
| 635 | 1479102518.248 | 1000     |       12 | complete      |
+-----+----------------+----------+----------+---------------+

预期结果：

+------------------------+----------+----------+----------+
| Ranges                 | Count    | Delta    |    %     | 
+------------------------+----------+----------+----------+
| Between 0 to 10 secs   |       44 | +44      |  84.62 % |
| Between 0 to 20 secs   |       48 | +4       |  92.31 % |
| Between 0 to 30 secs   |       50 | +2       |  96.15 % |
| Between 0 to 40 secs   |       51 | +1       |  98.08 % |
| Between 0 to 50 secs   |       51 | +0       |  98.08 % |
| Between 0 to 60 secs   |       51 | +0       |  98.08 % |
| Between 0 to 70 secs   |       51 | +0       |  98.08 % |
| Between 0 to 80 secs   |       52 | +1       | 100.00 % |
| Between 0 to 90 secs   |       52 | +0       | 100.00 % |
| Between 0 to 100+ secs |       52 | +0       | 100.00 % |
+------------------------+----------+----------+----------+
Total                            52

我现在能够创建的是以下查询，如果您可以帮助提供更好的解决方案，请提出建议。我现在面临的问题是（优先级）我无法获得百分比的计数和（次要）最终计数总数（52），现在我手动输入最终计数总数（52）。请帮忙。

SELECT Ranges,Delta,ROUND(Delta/52*100,2) AS '%'
FROM
(
    SELECT 
    (
        IF(duration<=10,'10',IF(duration<=20,'20',IF(duration<=30,'30',
        IF(duration<=40,'40',IF(duration<=50,'50',
        IF(duration<=60,'60',IF(duration<=70,'70',IF(duration<=80,
        '80',IF(duration<=90,'90','100+')))))))))) 
        AS Ranges,COUNT(duration) AS Delta
        FROM callsdetails
        GROUP BY Ranges
    ) a
GROUP BY Ranges;

当前结果：

+--------+-------+-------+
| Ranges | Delta | %     |
+--------+-------+-------+
| 10     |    44 | 84.62 |
| 20     |     4 |  7.69 |
| 30     |     2 |  3.85 |
| 40     |     1 |  1.92 |
| 80     |     1 |  1.92 |
+--------+-------+-------+

================================================ ==============================

感谢 Etsa 帮助我解决了上述问题，即所有呼叫的总数，现在我可能需要再次帮助，为每个队列名称添加分组。它与上面的公式相同，但现在必须按队列名称拆分或分组。

这是您可以使用数据进行测试的 sqlfiddle：http://sqlfiddle.com/#!9/efe11b/2

预期结果：

+---------+------+-----------+-------+--------+
| QUEUE   | RAN  | TOT_COUNT | DELTA | %      |
+---------+------+-----------+-------+--------+
| sales   | 10   |         6 |     6 | 100.00 |
| sales   | 20   |         6 |     0 | 100.00 |
| sales   | 30   |         6 |     0 | 100.00 |
| sales   | 40   |         6 |     0 | 100.00 |
| sales   | 50   |         6 |     0 | 100.00 |
| sales   | 60   |         6 |     0 | 100.00 |
| sales   | 70   |         6 |     0 | 100.00 |
| sales   | 80   |         6 |     0 | 100.00 |
| sales   | 90   |         6 |     0 | 100.00 |
| sales   | 100+ |         6 |     0 | 100.00 |
| enquiry | 10   |        32 |    32 |  80.00 |
| enquiry | 20   |        36 |     4 |  90.00 |
| enquiry | 30   |        38 |     2 |  95.00 |
| enquiry | 40   |        39 |     1 |  97.50 |
| enquiry | 50   |        39 |     0 |  97.50 |
| enquiry | 60   |        39 |     0 |  97.50 |
| enquiry | 70   |        39 |     0 |  97.50 |
| enquiry | 80   |        40 |     1 | 100.00 |
| enquiry | 90   |        40 |     0 | 100.00 |
| enquiry | 100+ |        40 |     0 | 100.00 |
| others  | 10   |         6 |     6 | 100.00 |
| others  | 20   |         6 |     0 | 100.00 |
| others  | 30   |         6 |     0 | 100.00 |
| others  | 40   |         6 |     0 | 100.00 |
| others  | 50   |         6 |     0 | 100.00 |
| others  | 60   |         6 |     0 | 100.00 |
| others  | 70   |         6 |     0 | 100.00 |
| others  | 80   |         6 |     0 | 100.00 |
| other   | 90   |         6 |     0 | 100.00 |
| others  | 100+ |         6 |     0 | 100.00 |
+---------+------+-----------+-------+--------+

【问题讨论】：

你试过了吗？ SO 不是免费软件服务。尝试做出努力，将其发布在您的问题中，然后寻求帮助。此外，我建议创建一个新问题。
嗨，Etsa，我真的很抱歉让我看起来像是在利用你。我一直在尝试多次修改您的代码，但我尝试的次数越多，显示的最终结果就越错误。非常抱歉打扰您，因为我对此不太熟悉，希望能得到您的帮助。真的很抱歉，感谢您的帮助。

标签： mysql group-by count range intervals

【解决方案1】：

一种方法可能是这样（使用变量来获得累积总和，并使用 JOIN 与所有间隔。最后我添加了一个带有 SELECT COUNT(*) 的 JOIN 来获得 52）：

SELECT RAN, TOT_COUNT, DELTA, ROUND(TOT_COUNT/E.TOT*100,2) AS '%'
FROM (SELECT B.RAN, a.Ranges, COALESCE(a.Delta,0) AS DELTA
      , @r:=@r+COALESCE(a.Delta,0) AS TOT_COUNT
      FROM (SELECT (
                    IF(duration<=10,'10',IF(duration<=20,'20',IF(duration<=30,'30',
                    IF(duration<=40,'40',IF(duration<=50,'50',
                    IF(duration<=60,'60',IF(duration<=70,'70',IF(duration<=80,
                    '80',IF(duration<=90,'90','100+')))))))))) 
                    AS Ranges,COUNT(duration) AS Delta
            FROM callsdetails C
            GROUP BY Ranges) A
        RIGHT JOIN (SELECT '10' AS RAN, 1 AS ORD UNION ALL SELECT '20', 2 AS ORD UNION ALL SELECT '30', 3
                UNION ALL SELECT '40', 4 UNION ALL SELECT '50', 5 UNION ALL SELECT '60',6 UNION ALL SELECT '70',7 
                UNION ALL SELECT '80',8 UNION ALL SELECT '90',9 UNION ALL SELECT '100+',10) B ON A.Ranges=B.RAN
        CROSS JOIN (SELECT @r:=0) T
        ORDER BY ORD
      ) D
CROSS JOIN (SELECT COUNT(*) AS TOT FROM callsdetails) E 
;

输出：

RAN TOT_COUNT   DELTA  %
10      44      44  84.62
20      48       4  92.31
30      50       2  96.15
40      51       1  98.08
50      51       0  98.08
60      51       0  98.08
70      51       0  98.08
80      52       1  100
90      52       0  100
100+    52       0  100

已更新（有问题的新请求）。

SELECT queueALL, RAN, TOT_COUNT, DELTA, ROUND(TOT_COUNT/D.TOT*100,2) AS '%'
FROM (
     SELECT B.queueALL, B.RAN, a.Ranges, COALESCE(a.Delta,0) AS DELTA, B.TOT
      , CASE WHEN queueALL=@q THEN @r:=@r+COALESCE(a.Delta,0) ELSE @r:=COALESCE(a.Delta,0) END AS TOT_COUNT
      , @q:=queueALL 
      , ORD
      FROM (SELECT queuename
             , CASE WHEN duration DIV 10.0001+1<=9 THEN CAST((duration DIV 10.0001+1)*10 AS CHAR) 
               ELSE '100+' END AS Ranges
             , COUNT(duration) AS Delta
            FROM callsdetails C
            GROUP BY queuename, Ranges) A
      RIGHT JOIN (SELECT B2.queuename AS queueALL, B2.TOT, B1.RAN, B1.ORD
                  FROM
                     (SELECT queuename, COUNT(*) AS TOT FROM callsdetails GROUP BY queuename) B2
                      CROSS JOIN (SELECT '10' AS RAN,1 AS ORD UNION ALL SELECT '20',2 AS ORD UNION ALL SELECT '30',3
                            UNION ALL SELECT '40',4 UNION ALL SELECT '50',5 UNION ALL SELECT '60',6 UNION ALL SELECT '70',7 
                            UNION ALL SELECT '80',8 UNION ALL SELECT '90',9 UNION ALL SELECT '100+',10)B1 ) B ON A.Ranges=B.RAN AND A.queuename = B.queueALL
      CROSS JOIN (SELECT @r:=0, @q:='') T
      ORDER BY queueALL, ORD) D
 ORDER BY queueALL, ORD

;

输出：

+----------+------+-----------+-------+------+
| queueALL | RAN  | TOT_COUNT | DELTA |  %   |
+----------+------+-----------+-------+------+
| enquiry  | 10   |        32 |    32 |   80 |
| enquiry  | 20   |        36 |     4 |   90 |
| enquiry  | 30   |        38 |     2 |   95 |
| enquiry  | 40   |        39 |     1 | 97.5 |
| enquiry  | 50   |        39 |     0 | 97.5 |
| enquiry  | 60   |        39 |     0 | 97.5 |
| enquiry  | 70   |        39 |     0 | 97.5 |
| enquiry  | 80   |        40 |     1 |  100 |
| enquiry  | 90   |        40 |     0 |  100 |
| enquiry  | 100+ |        40 |     0 |  100 |
| others   | 10   |         6 |     6 |  100 |
| others   | 20   |         6 |     0 |  100 |
| others   | 30   |         6 |     0 |  100 |
| others   | 40   |         6 |     0 |  100 |
| others   | 50   |         6 |     0 |  100 |
| others   | 60   |         6 |     0 |  100 |
| others   | 70   |         6 |     0 |  100 |
| others   | 80   |         6 |     0 |  100 |
| others   | 90   |         6 |     0 |  100 |
| others   | 100+ |         6 |     0 |  100 |
| sales    | 10   |         6 |     6 |  100 |
| sales    | 20   |         6 |     0 |  100 |
| sales    | 30   |         6 |     0 |  100 |
| sales    | 40   |         6 |     0 |  100 |
| sales    | 50   |         6 |     0 |  100 |
| sales    | 60   |         6 |     0 |  100 |
| sales    | 70   |         6 |     0 |  100 |
| sales    | 80   |         6 |     0 |  100 |
| sales    | 90   |         6 |     0 |  100 |
| sales    | 100+ |         6 |     0 |  100 |
+----------+------+-----------+-------+------+

【讨论】：

您好，Etsa，非常感谢您的解决方案，我遇到这个问题已经有一段时间了，非常感谢您的帮助。 (^_^)
您好，Etsa，我有新问题需要您的帮助。上次您帮助我解决了上述问题，即所有呼叫的总数，现在我将再次需要帮助，添加每个队列名称的分组。它与上面的公式相同，但现在必须按队列名称拆分或分组。请参阅上面的预期结果。这是您可以使用数据进行测试的 sqlfiddle：sqlfiddle.com/#!9/efe11b/2
@DarkSilver 添加了新版本，其中包括队列名称。
哇，你太擅长了，真的非常非常非常感谢你的帮助。我会从你的代码中学习和学习。 (^_^)