| summary.hmmmfit {hmmm} | R Documentation |
summary for the class hmmmfit
Description
The generic function ‘summary’ is adapted to the objects inheriting from class hmmmfit
(summary.hmmmfit) to display the results of the estimation of a hmm model by ‘hmmm.mlfit’.
Usage
## S3 method for class 'hmmmfit'
summary(object, cell.stats = TRUE, ...)Arguments
object |
An object of the class |
cell.stats |
If TRUE cell-specific statistics are returned |
... |
Further arguments passed to or from other methods |
Details
The marginal interactions of a hmm model can be defined in terms of linear predictor of covariates Cln(Mm)=Xbeta, where the X matrix is specified by ‘create.XMAT’ and the parameters beta indicate the additive effects of covariate on the marginal interactions. The function ‘hmmm.mlfit’ estimates either the parameters beta and the interactions; the function ‘summary’ of a fitted model (by ‘hmmm.mlfit’) returns the estimated betas and the estimated interactions, while the function ‘print’ provides the estimated interactions only. If the model is defined under equality constraints ECln(Mm)=0, parameters betas are meaningless so they are not printed.
The printed output of ‘summary’ provides: 1. values of the likelihood ratio and Pearson's score statistics, degrees of freedom and pvalues. Note that degrees of freedom and pvalues are meaningful only for the hmm models without inequality constraints (see ‘hmmm.chibar’ to test hmm models defined under inequality constraints on interactions); 2. the linear predictor model results: estimated betas, standard errors, z-ratios, pvalues; estimated interactions, standard errors, residuals; 3. cell-specific statistics: observed and predicted frequencies of the multi-way table, estimated joint probabilities with standard errors, adjusted residuals; 4. convergence statistics.
Value
No return value
Note
Use ‘print’ to display only the goodness-of-fit test and the estimated interactions.
See Also
hmmm.mlfit, print.hmmmfit, anova.hmmmfit, create.XMAT
Examples
data(relpol)
y<-getnames(relpol,st=12,sep=";")
# 1 = Religion, 2 = Politics
names<-c("Rel","Pol")
marglist<-c("l-m","m-g","l-g")
marginals<-marg.list(marglist,mflag="m")
# Hypothesis of stochastic independence: all log odds ratios are null
model<-hmmm.model(marg=marginals,lev=c(3,7),sel=c(9:20),names=names)
fitmodel<-hmmm.mlfit(y,model)
# print(fitmodel,aname="Independence model",printflag=TRUE)
summary(fitmodel)