sql解惑谜题3 麻醉师谜题
DROP TABLE PROCS;
CREATE TABLE PROCS AS
SELECT 10 AS proc_id, 'Baker' AS anest_name, '08:00' AS start_time, '11:00' AS end_time FROM dual UNION ALL
SELECT 20 AS proc_id, 'Baker' AS anest_name, '09:00' AS start_time, '13:00' AS end_time FROM dual UNION ALL
SELECT 30 AS proc_id, 'Dow' AS anest_name, '09:00' AS start_time, '15:30' AS end_time FROM dual UNION ALL
SELECT 40 AS proc_id, 'Dow' AS anest_name, '08:00' AS start_time, '13:30' AS end_time FROM dual UNION ALL
SELECT 50 AS proc_id, 'Dow' AS anest_name, '10:00' AS start_time, '11:30' AS end_time FROM dual UNION ALL
SELECT 60 AS proc_id, 'Dow' AS anest_name, '12:30' AS start_time, '13:30' AS end_time FROM dual UNION ALL
SELECT 70 AS proc_id, 'Dow' AS anest_name, '13:30' AS start_time, '14:30' AS end_time FROM dual UNION ALL
SELECT 80 AS proc_id, 'Dow' AS anest_name, '18:00' AS start_time, '19:00' AS end_time FROM dual;
我们首先生成各时间点
SELECT anest_name, start_time AS t FROM procs
UNION
SELECT anest_name, end_time AS t FROM procs;
ANEST_NAME T
---------- -----
Baker 08:00
Baker 09:00
Baker 11:00
Baker 13:00
Dow 08:00
Dow 09:00
Dow 10:00
Dow 11:30
Dow 12:30
Dow 13:30
Dow 14:30
Dow 15:30
Dow 18:00
Dow 19:00
14 rows selected
对于各时间点，分别取出对应的麻醉过程数据。
SELECT a.t, b.*, COUNT(*) over(PARTITION BY a.anest_name, a.t) AS ct
FROM (SELECT anest_name, start_time AS t FROM procs
UNION
SELECT anest_name, end_time AS t FROM procs) a
INNER JOIN procs b ON (b.anest_name = a.anest_name AND b.start_time <= a.t AND b.end_time > a.t)
ORDER BY 3, 1, 2;
T PROC_ID ANEST_NAME START_TIME END_TIME CT
----- ---------- ---------- ---------- -------- ----------
08:00 10 Baker 08:00 11:00 1
09:00 10 Baker 08:00 11:00 2
09:00 20 Baker 09:00 13:00 2
11:00 20 Baker 09:00 13:00 1
08:00 40 Dow 08:00 13:30 1
09:00 30 Dow 09:00 15:30 2
09:00 40 Dow 08:00 13:30 2
10:00 30 Dow 09:00 15:30 3
10:00 40 Dow 08:00 13:30 3
10:00 50 Dow 10:00 11:30 3
11:30 30 Dow 09:00 15:30 2
11:30 40 Dow 08:00 13:30 2
12:30 30 Dow 09:00 15:30 3
12:30 40 Dow 08:00 13:30 3
12:30 60 Dow 12:30 13:30 3
13:30 30 Dow 09:00 15:30 2
13:30 70 Dow 13:30 14:30 2
14:30 30 Dow 09:00 15:30 1
18:00 80 Dow 18:00 19:00 1
19 rows selected
然后分别取各麻醉过程最大重叠时间点即可
SELECT proc_id,anest_name, start_time,end_time, MAX(ct) AS ct
FROM
(
SELECT a.t, b.*, COUNT(*) over(PARTITION BY a.anest_name, a.t) AS ct
FROM (SELECT anest_name, start_time AS t
FROM procs
UNION
SELECT anest_name, end_time AS t
FROM procs) a
INNER JOIN procs b
ON (b.anest_name = a.anest_name AND b.start_time <= a.t AND
b.end_time > a.t)
)
GROUP BY proc_id,anest_name, start_time,end_time
ORDER BY 2,1;
PROC_ID ANEST_NAME START_TIME END_TIME CT
---------- ---------- ---------- -------- ----------
10 Baker 08:00 11:00 2
20 Baker 09:00 13:00 2
30 Dow 09:00 15:30 3
40 Dow 08:00 13:30 3
50 Dow 10:00 11:30 3
60 Dow 12:30 13:30 3
70 Dow 13:30 14:30 2
80 Dow 18:00 19:00 1
8 rows selected