计算每种类型每天插入的行数

Question

我需要创建一个查询，该查询将显示按类型排序的每天插入的查询量。这听起来很模糊，所以我将在下面演示：

+----+------+----------+
| id | type | inserted |
+----+------+----------+
|  1 | a    | 1/2/2017 |
|  2 | a    | 1/2/2017 |
|  3 | a    | 2/2/2017 |
|  4 | b    | 1/2/2017 |
|  5 | b    | 1/2/2017 |
|  6 | b    | 2/2/2017 |
|  7 | b    | 3/2/2017 |
|  8 | b    | 3/2/2017 |
+----+------+----------+

结果必须是：

+------+----------+--------+
| type |   date   | amount |
+------+----------+--------+
| a    | 1/2/2017 |      2 |
| a    | 2/2/2017 |      1 |
| b    | 1/2/2017 |      2 |
| b    | 2/2/2017 |      1 |
| b    | 3/2/2017 |      2 |
+------+----------+--------+

我已经尝试过使用GROUP BY和DISTINCT的一些查询，但我无法弄清楚如何不重复。

我已经尝试过了： SELECT date, type, count(*) FROM table GROUP BY date, type但是这给了我重复的日期/类型组合。这就是它的回报：

Answer 1

按type分组，按天分组（即TRUNC( inserted )）：

Oracle安装程序：

CREATE TABLE your_table ( id, type, inserted ) AS
  SELECT 1, 'a', TO_DATE( '2017-02-01 00:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL UNION ALL
  SELECT 2, 'a', TO_DATE( '2017-02-01 01:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL UNION ALL
  SELECT 3, 'a', TO_DATE( '2017-02-02 00:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL UNION ALL
  SELECT 4, 'b', TO_DATE( '2017-02-01 00:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL UNION ALL
  SELECT 5, 'b', TO_DATE( '2017-02-01 01:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL UNION ALL
  SELECT 6, 'b', TO_DATE( '2017-02-02 01:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL UNION ALL
  SELECT 7, 'b', TO_DATE( '2017-02-03 02:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL UNION ALL
  SELECT 8, 'b', TO_DATE( '2017-02-03 03:00:00', 'YYYY-MM-DD HH24:MI:SS' ) FROM DUAL;

<强>查询：

SELECT type,
       TRUNC( inserted ) AS "date",
       COUNT(1) AS amount
FROM   your_table
GROUP BY type,
      TRUNC( inserted );

<强>输出：

TYPE date                AMOUNT
---- ------------------- ------
a    2017-02-01 00:00:00      2
a    2017-02-02 00:00:00      1
b    2017-02-01 00:00:00      2
b    2017-02-02 00:00:00      1
b    2017-02-03 00:00:00      2

Answer 2

这对你有用吗？

select type, inserted,  count(inserted)
  from tmp_table 
  group by type, inserted