[R] Extracting individual parameter estimates from mmlcr
Rben
ben at mertonfolio.com
Wed Feb 24 01:02:00 CET 2010
I am new to mmlcr and am working on a latent class mixture model attempting
to identify
the trajectory and number of classes that best describes my data. I am able
to find model
parameters such as degrees of freedom, loglikelihood, and BIC.
For example, here is a cubic 3-class model I am using.
mmlcr1 <- mmlcr(outer = ~ 1 | ID, components = list(list(formula = Score ~
poly(wave,3),
class = "cnormlong", min = 1, max = 4)), data = datlong, n.groups = 3)
In this model I can find these parameters as follows:
mmlcrl$df
mmlcr1$loglikelihood
mmlcr1$BIC
And I can identify individual parameter estimates such as posterior
probabilities of
belonging to specific classes, specific data points, corresponding fitted
and residual values
as follows:
#posterior probability of belonging to class 1
summary.mmlcr(mmlcr1)[7]$post.prob$PostProb1
#first data value
summary.mmlcr(mmlcr1)$components[[1]]$data[1,]$PDS
#first fitted value
summary.mmlcr(mmlcr1)$components[[1]]$fitted[1]
#first residual value
summary.mmlcr(mmlcr1)$components[[1]]$residual[1]
BUT I am unable to figure out how to extract individual parameter estimates
such as intercepts, linear/quadratic/cubic slopes of my model. Is this found
in a similar way or can it be computed. If it can be computed, does anyone
know
how this is done in the context of this program?
Thanks for any help.
--
View this message in context: http://n4.nabble.com/Extracting-individual-parameter-estimates-from-mmlcr-tp1566791p1566791.html
Sent from the R help mailing list archive at Nabble.com.
More information about the R-help
mailing list