envelope {PLreg}

Normal Probability Plots with Simulated Envelope of Residuals for PLreg Objects

Description

envelope is used to display normal probability plots with simulated envelope of residuals for the power logit models. Currently, three types of residuals are supported: quantile, deviance and standardized residuals.

Usage

envelope(
  object,
  type = c("quantile", "deviance", "standardized"),
  rep = 40,
  conf = 0.95,
  xlab,
  ylab,
  main,
  envcol,
  ylim,
  xlim
)

Arguments

object

fitted model object of class "PLreg".

type

character specifying the type of residuals to be used, see residuals.PLreg. Default is type = "standardized".

rep

a positive integer representing the number of iterations to calculate the simulated envelopes. Default is rep=40.

conf

a numeric value in the interval (0,1) that represents the confidence level of the simulated envelopes. Default is conf=0.95.

xlab

character specifying the label for x axis (optional). Default is "Quantile N(0,1)".

ylab

character specifying the label for y axis (optional). Default is the name of the used residual.

main

character specifying the overall title for the plot.

envcol

character specifying the color of the envelope.

ylim, xlim

numeric values, specifying the left/lower limit and the right/upper limit of the scale.

Details

The envelope uses the idea of Atkinson (1985) to create normal probability plots with simulated envelope. Under the correct model, approximately 100*conf of the residuals are expected to be inside the envelope.

Value

envelope returns normal probability plot with simulated envelopes for the residuals.

References

Queiroz, F. F. and Ferrari, S. L. P. (2022). Power logit regression for modeling bounded data. arXiv:2202.01697.

Atkinson, A. C. (1985) Plots, transformations and regression: an introduction to graphical methods of diagnostic regression analysis. Oxford Science Publications, Oxford.

See Also

PLreg, residuals.PLreg

Examples

data("Firm")

fitPL <- PLreg(firmcost ~ sizelog + indcost | sizelog + indcost, data = Firm,
              family = "SLASH", zeta = 2.13)
summary(fitPL)

envelope(fitPL, type = "standardized")
envelope(fitPL, type = "quantile")
envelope(fitPL, type = "deviance")