piplot.ct: Pre-analysis Plot for 'pistar.ct'

In jmedzihorsky/pistar: Rudas, Clogg and Lindsay Mixture Index of Fit

Pre-analysis Plot for pistar.ct

Description

piplot.ct is used to inform the choice of the interval on which to look for pi* with pistar.ct. It plots log-likelihood ratio statistic values of a two-point mixture of a user-supplied model and an unrestricted component fit to a contingency table on a specified number of equally-spaced points on a supplied interval.

Usage

piplot.ct(fn, data, ..., from = .Machine$double.neg.eps^0.25,
          to = 1 - .Machine$double.neg.eps^0.25,  n = 10, draw = TRUE, 
          color = "black", zero_line = TRUE, add = FALSE, values = FALSE)

Arguments

fn	a user-supplied function that inputs a contingency table of observed values and outputs a contingency table of predicted values. The function must output a named list with the contingency table of predicted values named 'fit'.
data	a contingency table.
...	further arguments passed to the user-supplied function.
from	numeric: lower bound of the interval of out-of-model proportions to be explored.
to	numeric: upper bound of the interval of out-of-model proportions to be explored.
n	numeric: number of equally-spaced points from the interval of out-of-model proportions to be explored.
draw	logical: draw a plot or lines?
color	color of the line.
zero_line	logical: plot a horizontal line at 0?
add	logical: add the line to a plot?
values	logical: return a dataframe with explored out-of-model proportions and their corresponding log-likelihood ratio statistics?

Details

Developed from John M. Grego's clr.plot

Value

A plot. If values = TRUE returns also an object of class 'PiplotCT' with the following slots:

pi	numeric vector of explored out-of-model proportions
lr	numeric vector of corresponding log-likelihood ratio statistics
lr_plus_eps	lr plus the value of eps

Author(s)

Juraj Medzihorsky, developed from John M. Grego's clr.plot

References

Grego, J.M. clr.plot function available at http://www.stat.sc.edu/~grego/courses/stat770/CLR.txt

See Also

pistar.ct rcl.em

Examples

	data(Fienberg1980a)

	mf <- function(data) loglin(data, list(1, 2), fit=TRUE, print=FALSE)

	a <- piplot.ct(fn=mf, data=Fienberg1980a, values=TRUE)

	a

	plot(a, color="red")

jmedzihorsky/pistar documentation built on June 4, 2022, 9:58 a.m.