|
|---|
|
|
|
|
|
|
s**y
发帖数: 358 | 1 有两个table：
Table.1 接种的subject和接种时间：
subject Time
1 01/08/2014 08:21:00 AM
1 01/08/2014 11:44:00 AM
2 01/11/2014 04:57:00 AM
2 01/11/2014 09:24:00 AM
3 01/07/2014 03:41:00 PM
3 01/07/2014 04:52:00 PM
4 01/10/2014 03:46:00 PM
4 01/10/2014 05:15:00 PM
5 01/12/2014 04:59:00 AM
5 01/12/2014 04:59:00 AM
6 01/07/2014 10:39:00 AM
6 01/07/2014 01:07:00 PM
6 01/09/2014 02:09:00 PM
7 01/09/2014 02:37:00 PM
7 01/09/2014 04:30:00 PM
Table.2 这些subject接种前后一个礼拜的某项指标测量数据：
Subject Measure Time
1 10.50 01/07/2014 02:27:26 PM
1 7.10 01/08/2014 06:41:49 AM
1 9.10 01/09/2014 08:31:49 AM
2 11.60 01/08/2014 10:49:42 PM
2 10.20 01/09/2014 04:00:23 AM
2 8.70 01/10/2014 05:26:10 AM
2 6.70 01/11/2014 02:04:30 AM
2 10.30 01/11/2014 07:17:34 PM
3 9.80 01/05/2014 04:21:41 AM
3 9.20 01/06/2014 05:40:17 AM
3 7.80 01/06/2014 08:45:51 PM
3 7.40 01/07/2014 03:47:08 AM
3 6.70 01/07/2014 09:19:03 AM
3 9.70 01/08/2014 10:29:14 AM
3 9.80 01/07/2014 11:23:25 PM
3 10.10 01/08/2014 06:28:33 PM
3 9.60 01/09/2014 05:54:28 AM
3 10.10 01/09/2014 12:45:29 PM
3 8.90 01/10/2014 03:03:18 AM
3 9.70 01/10/2014 12:28:25 PM
3 10.00 01/11/2014 06:27:15 AM
3 9.10 01/11/2014 03:15:56 PM
4 9.10 01/10/2014 01:59:23 PM
4 8.00 01/10/2014 02:37:28 PM
4 10.80 01/10/2014 09:35:21 PM
4 9.50 01/11/2014 05:07:49 AM
4 10.00 01/11/2014 10:41:21 AM
4 9.20 01/11/2014 05:42:13 PM
5 11.50 01/11/2014 10:20:28 AM
6 14.20 01/05/2014 03:14:44 PM
6 12.30 01/06/2014 12:51:00 PM
6 8.20 01/07/2014 04:21:11 AM
6 7.00 01/07/2014 09:18:20 AM
6 8.90 01/07/2014 03:10:34 PM
6 9.00 01/07/2014 06:55:55 PM
6 8.70 01/07/2014 10:24:29 PM
6 8.20 01/08/2014 02:43:40 AM
6 8.70 01/08/2014 12:14:16 PM
6 7.00 01/09/2014 05:09:55 AM
6 8.60 01/10/2014 06:20:07 AM
6 8.00 01/11/2014 07:04:32 AM
7 10.10 01/06/2014 12:52:40 AM
7 7.80 01/07/2014 06:18:49 AM
7 5.10 01/09/2014 11:41:13 AM
7 4.90 01/09/2014 01:34:35 PM
7 7.80 01/10/2014 01:45:15 AM
7 7.50 01/11/2014 06:05:22 AM
问题： 如何找到这些subject：
1、接种前的最后两次measurement均值>7;
2、接种后的头两次measurement均值>7;
3.接种前后两次的measurement相差<1;
非常感谢！！ | s**y
发帖数: 358 | 2 这是目前想到的方法，但觉得不够精炼，有没有更好的办法？
求大牛出来指点一下！
data one;
input @1 subject @5 date mdyampm25.2;
format date mdyampm25.2;
datalines;
1 01/08/2014 08:21:00 AM
1 01/08/2014 11:44:00 AM
2 01/11/2014 04:57:00 AM
2 01/11/2014 09:24:00 AM
3 01/07/2014 03:41:00 PM
3 01/07/2014 04:52:00 PM
...
;
format mdyampm25.2;
run;
data two;
input Subject Measure @13 Time mdyampm25.2;
format time mdyampm25.2;
datalines;
Subject Measure Time
1 10.50 01/07/2014 02:27:26 PM
1 7.10 01/08/2014 06:41:49 AM
1 9.10 01/09/2014 08:31:49 AM
2 11.60 01/08/2014 10:49:42 PM
2 10.20 01/09/2014 04:00:23 AM
...
;
run;
data shot;
set one;
by subject date;
if first.subject &first.date then id=1;
else id+1;
run;
proc transpose data=shot out=shot1 prefix=time;
by subject;
var date;
id id;
run;
proc sort data=two ;
by subject descending time ;
run;
proc sql;
select a.subject, a.time, a.measure from two as a inner join shot1 as b on a
.subject= b.subject&a.time
where (select count(*) from two a2 where a2.time<=a.time and a2.subject=a.
subject)<=2
group by a.subject
having avg(measure)>7;
quit; |
|
|
|
|
|
|
