Model Diagnostic Plots for QRR-VGLMs

The residuals of a QRR-VGLM are plotted for model diagnostic purposes.

Usage

plotqrrvglm(object, rtype = c("response", "pearson", "deviance", "working"),
            ask = FALSE,
            main = paste(Rtype, "residuals vs latent variable(s)"),
            xlab = "Latent Variable",
            I.tolerances = object@control$eq.tolerances, ...)

Arguments

object: An object of class "qrrvglm".
rtype: Character string giving residual type. By default, the first one is chosen.
ask: Logical. If TRUE, the user is asked to hit the return key for the next plot.
main: Character string giving the title of the plot.
xlab: Character string giving the x-axis caption.
I.tolerances: Logical. This argument is fed into Coef(object, I.tolerances = I.tolerances).
...: Other plotting arguments (see par).

Details

Plotting the residuals can be potentially very useful for checking that the model fit is adequate.

Value

The original object.

References

Yee, T. W. (2004). A new technique for maximum-likelihood canonical Gaussian ordination. Ecological Monographs, 74, 685–701.

Author

Thomas W. Yee

Note

An ordination plot of a QRR-VGLM can be obtained by lvplot.qrrvglm.

Examples

if (FALSE) { # \dontrun{
# QRR-VGLM on the hunting spiders data
# This is computationally expensive
set.seed(111)  # This leads to the global solution
hspider[, 1:6] <- scale(hspider[, 1:6])  # Standardize environ vars
p1 <- cqo(cbind(Alopacce, Alopcune, Alopfabr, Arctlute, Arctperi,
                Auloalbi, Pardlugu, Pardmont, Pardnigr, Pardpull,
                Trocterr, Zoraspin) ~
          WaterCon + BareSand + FallTwig + CoveMoss + CoveHerb + ReflLux,
          poissonff, data = hspider, Crow1positive = FALSE)
par(mfrow = c(3, 4))
plot(p1, rtype = "response", col = "blue", pch = 4, las = 1, main = "")
} # }