Dear All,
I am trying to estimate a mixed effect model with random slope with
npmlreg. To make my question clear, I use the sample data set that was
used in the vignettes (as part of the package "nlme").
-----------------------------------------------------------------------------------------------------------------------------------------
> vc2 <- allvc(height ~ age, random=~age|Subject, data=Oxboys,
> random.distribution="np", k=3)
1 ..2 ..3 ..4 ..5 ..6 ..7 ..8 ..9 ..10 ..
EM algorithm met convergence criteria at iteration # 10
Disparity trend plotted.
EM Trajectories plotted.
> summary(vc2)
Call: allvc(formula = height ~ age, random = ~age | Subject, data =
Oxboys, k = 3, random.distribution = "np")
Coefficients:
Estimate Std. Error t value
age 7.919030 0.4065465 19.478782
MASS1 138.588240 0.2827517 490.141113
MASS2 149.249701 0.1921859 776.590184
MASS3 158.909797 0.2627202 604.863195
MASS1:age -2.350977 0.5966915 -3.940021
MASS2:age -1.701525 0.5034540 -3.379703
Mixture proportions:
MASS1 MASS2 MASS3
0.2313332 0.5007243 0.2679425
Component distribution - MLE of sigma: 3.586
Random effect distribution - standard deviation: 7.161265
-2 log L: 1315 Convergence at iteration 10
-----------------------------------------------------------------------------------------------------------------------------------------
My question is: how to interpret the the random slope coefficients
"age", "MASS1:age", and "MASS2:age"? Does it mean that the effect of
age is 7.919030 in the third component, -2.350977 in the first
component, and -1.701525 in the second, or something else?
Many thanks.
Best,
Shige
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
