SAS中文论坛

标题: 求助SAS做错位相关的问题！ [打印本页]

作者: shiyiming 时间: 2009-8-7 22:09
标题: 求助SAS做错位相关的问题！
数据是这样的：
date time a b c d e f
2007-9-20 15:50 43.26 31.4 906.1812 0.248863974 0.267639823 0.196451616
2007-9-20 16:00 45 31.25 803.6607 0.248863974 0.267369347 0.196401628
2007-9-20 16:10 44.54 31.15 732.6465 0.24857767 0.267369347 0.196265886
.......
.......
2007-9-24 23:20 88.17 25.77 17.5035 0.240853178 0.287935747 0.19089574
2007-9-24 23:30 88.37 25.77 17.5035 0.240853178 0.287935747 0.19089574
2007-9-24 23:40 88.68 25.77 17.0034 0.240853178 0.287935747 0.190939715
2007-9-24 23:50 90.37 25.67 17.0034 0.240853178 0.287935747 0.190939715
;
我需要将数据以date分组，两组数据之间进行错位相关分析（以10分钟为步长，每错一次位统计出相关系数），取相关系数最高时刻错位的时间差。因做毕业论文用，求高人指点，不甚感激！

作者: shiyiming 时间: 2009-8-10 13:36
标题: Re: 求助SAS做错位相关的问题！
用lag？

作者: shiyiming 时间: 2009-8-10 17:36
标题: Re: 求助SAS做错位相关的问题！
不明白楼主的要求，能否说得再清楚些？

作者: shiyiming 时间: 2009-8-11 08:17
标题: Re: 求助SAS做错位相关的问题！
简单点说，问题就在错位相关分析上。比如以下数据：
date time x1 x2
2008-10-30 0:00 2.15E-03 0.573645243
2008-10-30 0:10 0.301287953 0.573645243
2008-10-30 0:20 0.302631208 0.757863977
2008-10-30 0:30 0.496946252 0.757863977
2008-10-30 0:40 0.708597836 0.573645243
2008-10-30 0:50 0.93469088 0.757863977
2008-10-30 1:00 0.936369511 0.757863977
2008-10-30 1:10 1.175199353 0.952222615
2008-10-30 1:20 0.939728463 0.757863977
2008-10-30 1:30 0.941408781 0.757863977
2008-10-30 1:40 0.141656416 0.242603172
2008-10-30 1:50 0.142818888 0.103014422
2008-10-30 2:00 0 0
2008-10-30 2:10 0.317479989 0.242603172
2008-10-30 2:20 0.146316899 0.401075287
2008-10-30 2:30 0.147486405 0.103014422
2008-10-30 2:40 0.148657648 0.242603172
2008-10-30 2:50 0.322912237 0
2008-10-30 3:00 1.40E-02 5.33E-04
2008-10-30 3:10 0 0.106545246
2008-10-30 3:20 0 2.06E-03
2008-10-30 3:30 0 2.93E-03
2008-10-30 3:40 0 3.86E-03
2008-10-30 3:50 0 4.83E-03
2008-10-30 4:00 0 0
2008-10-30 4:10 0 0.117268061
2008-10-30 4:20 0 7.95E-03
2008-10-30 4:30 0 0.424034477
2008-10-30 4:40 0 0.122700156
2008-10-30 4:50 2.27E-02 0.267703387
2008-10-30 5:00 0.165232767 0.269815678
2008-10-30 5:10 0.538704149 0.433284374
2008-10-30 5:20 0.540207049 0.435602661
2008-10-30 5:30 0.541710731 0.799503258
2008-10-30 5:40 0.983597469 0.998519343
2008-10-30 5:50 1.475041638 1.001255408
2008-10-30 6:00 1.737637143 1.209513263
2008-10-30 6:10 2.578184658 1.876393253
2008-10-30 6:20 3.182654538 1.879512608
2008-10-30 6:30 3.18483191 1.653724683
2008-10-30 6:40 2.584416256 2.121662439
2008-10-30 6:50 2.883707461 1.888876477
2008-10-30 7:00 2.299814715 2.128068281
2008-10-30 7:10 1.750992881 1.895123885
2008-10-30 7:20 1.491636387 1.446804147
2008-10-30 7:30 1.002269121 1.235164683
2008-10-30 7:40 1.244150976 1.238021038
2008-10-30 7:50 1.245928571 1.034196617
2008-10-30 8:00 0.779081893 1.036950683
2008-10-30 8:10 0.780698177 1.039706117
2008-10-30 8:20 0.782315087 2.153727661
2008-10-30 8:30 1.253043694 4.275034965
2008-10-30 8:40 1.506416326 6.43098927
2008-10-30 8:50 2.611452274 9.592871395
2008-10-30 9:00 3.217525836 13.200551
2008-10-30 9:10 5.220934229 17.27752988
2008-10-30 9:20 7.097044378 21.32553555
2008-10-30 9:30 8.730765475 24.64531428
2008-10-30 9:40 11.40244167 29.47809085
2008-10-30 9:50 13.33656411 34.12437941
2008-10-30 10:00 15.3966938 36.97190014
2008-10-30 10:10 17.58759834 40.73773937
2008-10-30 10:20 18.73557466 43.11775051
2008-10-30 10:30 20.52149341 44.75843487
2008-10-30 10:40 20.52512377 44.76634717
2008-10-30 10:50 21.7620634 44.77425974
2008-10-30 11:00 21.76576828 44.78217256
2008-10-30 11:10 22.40000768 47.32000907
2008-10-30 11:20 23.04368301 48.19341277
2008-10-30 11:30 22.40749284 45.63855132
2008-10-30 11:40 21.15918505 39.27300887
2008-10-30 11:50 19.35284954 36.32791247
2008-10-30 12:00 18.7707798 38.53572
2008-10-30 12:10 18.77430099 34.92075509
2008-10-30 12:20 18.20082063 34.23029554
2008-10-30 12:30 18.78134375 38.55797075
2008-10-30 12:40 20.56869738 34.94211864
2008-10-30 12:50 18.21127589 33.56243913
2008-10-30 13:00 18.21476123 34.25855021
2008-10-30 13:10 17.08958358 33.57645251
2008-10-30 13:20 17.65311222 34.27267917
2008-10-30 13:30 17.65656158 36.40029933
2008-10-30 13:40 18.80599728 37.86650103
2008-10-30 13:50 17.66346068 32.92368541
2008-10-30 14:00 16.01127462 29.01988649
2008-10-30 14:10 12.92339624 25.398506
2008-10-30 14:20 12.43894327 24.8261737
2008-10-30 14:30 13.42504998 26.59046124
2008-10-30 14:40 13.9315024 27.80627283
2008-10-30 14:50 16.57168974 32.29274105
2008-10-30 15:00 17.6876115 34.34334034
2008-10-30 15:10 17.69106212 34.35040795
2008-10-30 15:20 16.03800952 30.34478259
2008-10-30 15:30 13.94749493 26.62886908
2008-10-30 15:40 11.98352728 23.73960471
2008-10-30 15:50 11.5140683 22.09359523
2008-10-30 16:00 10.14488482 21.56257077
2008-10-30 16:10 9.270330414 21.56850292
2008-10-30 16:20 10.60065638 22.11154581
2008-10-30 16:30 10.60360441 21.04996253
2008-10-30 16:40 10.15653162 20.53207204
2008-10-30 16:50 9.716851718 20.02069422
2008-10-30 17:00 7.617366417 17.53705516
2008-10-30 17:10 6.446401288 15.20873621
2008-10-30 17:20 5.341062982 13.02891362
2008-10-30 17:30 4.302115044 10.21411152
2008-10-30 17:40 3.022709063 8.369803142
2008-10-30 17:50 2.724313303 7.329517376
2008-10-30 18:00 1.875909719 6.664052365
2008-10-30 18:10 1.612500089 5.087741201
2008-10-30 18:20 1.614373168 4.208534055
2008-10-30 18:30 1.881713665 3.928952446
2008-10-30 18:40 2.442363544 3.116786589
2008-10-30 18:50 1.364279078 3.120262606

先将x1和x2做相关分析，返回一个相关系数，再将x2向下错位一行，空出的地方用.代替，再进行一次相关分析，返回相关系数，并和上个相关系数比较大小，如果相关系数比上一个大，则继续错位相关分析，如果比上一个小则停止，同时输出错位的时间差（即错位个数乘以10min）。不知道这么说是否详细？

作者: shiyiming 时间: 2009-8-11 14:28
标题: Re: 求助SAS做错位相关的问题！
做法不是很完美，但是可以解决问题。等待更好的方法：）
data aa;
input date $10. time time5. x1 x2;
cards;
2008-10-30 0:00 2.15E-03 0.573645243
.
.
.
.
.
.
.
;
run;

data bb;
set aa ;
keep x1;
run;

data bb;
set bb;
rename x1=x2;
run;

data cc;
set aa bb;
run;
quit;

175是总的观测数115+60的来，只需要在output中观察从116到175的自相关系数，找到到最大值。

proc arima data=cc ;
identify var=x2 nlag=175;
quit;

欢迎光临 SAS中文论坛 (https://mysas.net/forum/)