Date: Mon, 10 Mar 2008 04:46:17 -0700
Reply-To: Steve Denham <stevedrd@YAHOO.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Steve Denham <stevedrd@YAHOO.COM>
Subject: Re: LSMEAN Non-est
Content-Type: text/plain; charset=us-ascii
Kanimozhi,
Sequence should be nested within period by centre for this design. The following code works:
PROC GLM data=abc;
class rnid treat period sequence centre;
model diff = sequence rnid(sequence*centre) period treat centre/ SS3;
lsmeans treat /pdiff CL alpha=0.05 e;
estimate 'T VS R' treat 1 -1;
run;
quit;
Note that I also added the /e option to the lsmeans statement to check estimability. If you include it for your original model, you will find that rnid is completely confounded with sequence if you ignore centre. I think that's probably the source of the nonestimability.
Steve Denham
Associate Director, Biostatistics
MPI Research
----- Original Message ----
From: Kanimozhi A. <kanimozhi.a@ACUNOVALIFE.COM>
To: SAS-L@LISTSERV.UGA.EDU
Sent: Monday, March 10, 2008 6:13:18 AM
Subject: LSMEAN Non-est
Dear Members,
Following is the data and the code I used to estimate the LSMEAN and the
Confidence interval.
data abc;
input rnid 1-2 sequence 5-6 period 9-10 treat 13-14 centre 17-18 diff
22-32;
cards;
1 1 1 1 2 1.744708995
1 1 2 2 2 3.928571429
2 2 1 2 2 3.821428571
2 2 2 1 2 3.25
3 2 1 2 2 3.544973545
3 2 2 1 2 2.297619048
4 1 1 1 2 3.07010582
4 1 2 2 2 2.5
5 2 1 2 2 3.802910053
5 2 2 1 2 3.678571429
7 2 1 2 2 3.464285714
7 2 2 1 2 4.678571429
9 2 1 2 2 2.821428571
9 2 2 1 2 3.107142857
11 2 1 2 1 2.285714286
11 2 2 1 1 2.642857143
12 1 1 1 1 1.714285714
12 1 2 2 1 1
13 1 1 1 1 2.214285714
13 1 2 2 1 2.571428571
14 2 1 2 1 2.25
14 2 2 1 1 2.571428571
15 1 1 1 1 -0.33994709
15 1 2 2 1 1.107142857
17 2 1 2 1 0.571428571
17 2 2 1 1 1
18 1 1 1 1 1
18 1 2 2 1 3.107142857
19 2 1 2 1 2.357142857
19 2 2 1 1 3.321428571
20 1 1 1 1 1.321428571
20 1 2 2 1 1.678571429
21 2 1 2 2 3.321428571
21 2 2 1 2 2.857142857
22 1 1 1 2 3.642857143
22 1 2 2 2 4.035714286
23 2 1 2 2 2.214285714
23 2 2 1 2 3.321428571
24 1 1 1 2 1.464285714
24 1 2 2 2 3.943121693
26 2 1 2 2 2.5
26 2 2 1 2 3.616402116
28 1 1 1 2 1.892857143
28 1 2 2 2 2.079365079
29 2 1 2 2 1.25
29 2 2 1 2 2.5
30 1 1 1 2 2.642857143
30 1 2 2 2 3
31 1 1 1 3 0.25
31 1 2 2 3 1.821428571
32 2 1 2 3 1.428571429
32 2 2 1 3 2.357142857
33 1 1 1 3 0.714285714
33 1 2 2 3 0.607142857
34 1 1 1 3 0.642857143
34 1 2 2 3 2.25
35 2 1 2 3 0.75
35 2 2 1 3 0.071428571
38 2 1 2 1 -0.142857143
38 2 2 1 1 0.285714286
39 1 1 1 1 1.29945055
39 1 2 2 1 3.142857143
40 2 1 2 1 1.642857143
40 2 2 1 1 1.785714286
41 2 1 2 1 1.464285714
41 2 2 1 1 2.357142857
42 1 1 1 1 2.071428571
42 1 2 2 1 0.321428571
43 1 1 1 1 2
43 1 2 2 1 1.107142857
44 1 1 1 1 2
44 1 2 2 1 2.75
45 2 1 2 1 2.214285714
45 2 2 1 1 0.642857143
46 2 1 2 3 0.928571429
46 2 2 1 3 1.357142857
49 1 1 1 3 0.928571429
49 1 2 2 3 0.357142857
50 2 1 2 1 0.522486773
50 2 2 1 1 1.714285714
51 1 1 1 1 2.428571429
51 1 2 2 1 2.892857143
52 2 1 2 1 1.142857143
52 2 2 1 1 1.105820106
;
run;
PROC GLM data=abc;
class rnid treat period sequence centre;
model diff = sequence rnid(sequence) period treat centre/ SS3;
lsmeans treat /pdiff CL alpha=0.05;
estimate 'T VS R' treat 1 -1;
run;
quit;
by using the above code I could not estimate LSMEAN, in the output
window I got Non-est instead of the value.
Could explain me why I could not estimate, and how to correct it.
I would be grateful if I can be clarified.
Thanks in advance.
Regards,
Kanimozhi
!! Disclaimer Message !!
This e-mail message may contain privileged or confidential information intended solely for the use of the addressee(s). If you are not the intended recipient, you may not disclose, use, disseminate, distribute, copy or rely upon this message or attachment in any way. If you received this e-mail message in error, please return by forwarding the message and its attachments to the sender. This e-mail may contain viruses. Manipal AcuNova has taken every reasonable precaution to minimize this risk, but is not liable for any damage you may sustain as a result of any virus in this e-mail. You should carry out your own virus checks before opening the e-mail or attachment. Manipal AcuNova do not accept liability for any errors, omissions, corruption in the contents of this message or any attachments.
!! End Of Disclaimer Message !!
____________________________________________________________________________________
Looking for last minute shopping deals?
Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping
