fitted.eglhmm: Predict method for extended generalised linear hidden Markov...

In eglhmm: Extended Generalised Linear Hidden Markov Models

Predict method for extended generalised linear hidden Markov models.

Description

Predicted values based on an extended generalised linear hidden Markov model object.

Usage

## S3 method for class 'eglhmm'
fitted(object, ...)

Arguments

object	An object of class eglhmm as returned by eglhmm().
...	Not used.

Value

A vector of fitted values of the same length as that of the observed values (i.e. length equal to the row dimension of the data frame to which the model was fitted. This data frame is equal to object$data but with repeated rows corresponding to different states collapsed to a single row. The row dimension of this data frame is thus nrow(object$data)/K where K is the number of states in the model. This data frame, with columns cf and state omitted, is returned as an attribute data of the vector of fitted values.

Remark

Although this documentation refers to “generalised linear models”, the only such models currently (\today) available are the Gaussian model with the identity link, the Poisson model, with the log link, and the Binomial model with the logit link. Other models may be added at a future date.

Author(s)

Rolf Turner [e-mail]

References

See the help for eglhmm() for references.

See Also

reglhmm() reglhmm.default() reglhmm.eglhmm() bcov()

Examples

    loc4 <- c("LngRf","BondiE","BondiOff","MlbrOff")
    SCC4 <- SydColCount[SydColCount$locn %in% loc4,] 
    SCC4$locn <- factor(SCC4$locn) # Get rid of unused levels.
    rownames(SCC4) <- 1:nrow(SCC4)
    fit <- eglhmm(y~locn+depth,data=SCC4,cells=c("locn","depth"),
                 K=2,distr="P",contr="sum",verb=TRUE)
    fv  <- fitted(fit)
    with(attr(fv,"data"),plot(y[locn=="BondiOff" & depth=="40"],
             xlab="time",ylab="count"))
    with(attr(fv,"data"),lines(fv[locn=="BondiOff" & depth=="40"]))

eglhmm documentation built on May 29, 2024, 1:20 a.m.