Hi,
I am looking for help generating Lines for lsmeans in Proc GLIMMIX. I am providing my SAS code, sources of variance, and error. I am also including what I believe the probable reason for the error. I was wondering if there is any way around it. Disclaimer, I am not an advanced user of SAS.
SAS code as follows:
proc glimmix data = G5_H2;
class E G Y H R Count;
MODEL weight =
E
Y
G
E*G
E*G*Y
Y*G
H
H*E
H*Y
H*G
H*Y*G
H*E*G
H*E*Y
H*E*Y*G/ddfm = kenwardroger;
RANDOM
E*Y
E*G*R
E*Y*G*R
E*H*G*R
E*H*Y*G*R;
lsMEANS G/pdiff=all adjust=tukey pdiff=all lines;
lsMEANS G*E/ slice=G adjust=tukey pdiff=all lines;
lsMEANS G*Y/ slice=G adjust=tukey pdiff=all lines;
lsMEANS G*E*Y/ slice=G adjust=tukey pdiff=all lines;
*lsMEANS H*E*Y*G/ slice=G adjust=tukey;
run;
Sources:
E = 2
Y = 2
G = 5
H = 2
R = 2
Count = 10* for most of the samples see below
Note:
There is not a full 10 count for a few individuals in E=2, Y=2, and H=2 (colored orange in the diagram)
For one individual in G in E=2 and Y=1 there is only 1 R, but 2 H; further, for the same G in E=2 and Y=2 there are 2 R but only 1 H. See the following diagram (displayed in bold red).
E | Y | H | R | G1 | G2 | G3 | G4 | G5 |
1 | 1 | 1 | 1 | 10 | 10 | 10 | 10 | 10 |
1 | 1 | 1 | 2 | 10 | 10 | 10 | 10 | 10 |
1 | 1 | 2 | 1 | 10 | 10 | 10 | 10 | 10 |
1 | 1 | 2 | 2 | 10 | 10 | 10 | 10 | 10 |
1 | 2 | 1 | 1 | 10 | 10 | 10 | 10 | 10 |
1 | 2 | 1 | 2 | 10 | 10 | 10 | 10 | 10 |
1 | 2 | 2 | 1 | 10 | 10 | 10 | 10 | 10 |
1 | 2 | 2 | 2 | 10 | 10 | 10 | 10 | 10 |
2 | 1 | 1 | 1 | 10 | 10 | 10 | 0 | 10 |
2 | 1 | 1 | 2 | 10 | 10 | 10 | 10 | 10 |
2 | 1 | 2 | 1 | 10 | 10 | 10 | 0 | 10 |
2 | 1 | 2 | 2 | 10 | 10 | 10 | 10 | 10 |
2 | 2 | 1 | 1 | 10 | 10 | 10 | 10 | 10 |
2 | 2 | 1 | 2 | 10 | 10 | 10 | 10 | 10 |
2 | 2 | 2 | 1 | 10 | 7 | 8 | 0 | 10 |
2 | 2 | 2 | 2 | 10 | 5 | 10 | 0 | 7 |
The error code:
NOTE: With DDFM=SATTERTHWAITE or DDFM=KENWARDROGER, unadjusted p-values in tests are based on the
degrees of freedom specific to that comparison. P-values that are adjusted for multiplicity,
however, are by default based on the denominator degrees of freedom for the Type III test of
the fixed effect. If you specify the ADJDFE=ROW option in the LSMEANS or LSMESTIMATE statement,
the adjusted p-values take into account the row-wise degrees of freedom.
NOTE: Some observations are not used in the analysis because of: missing response values (n=3).
NOTE: Convergence criterion (GCONV=1E-8) satisfied.
NOTE: Estimated G matrix is not positive definite.
ERROR: LINES display is not produced for the G effect because adjusted p-values are not
available for all least-squares means differences.
ERROR: LINES display is not produced for the E*G effect because adjusted p-values
are not available for all least-squares means differences.
ERROR: LINES display is not produced for the G*Y effect because adjusted p-values are not
available for all least-squares means differences.
ERROR: LINES display is not produced for the E*G*Y effect because adjusted p-values
are not available for all least-squares means differences.
NOTE: PROCEDURE GLIMMIX used (Total process time):
real time 2.46 seconds
cpu time 2.42 seconds