pscl (version 0.5)

plot.ideal: plots an ideal object.

Description

plot of the results of an ideal point estimation contained in an object of class ideal

Usage

plot.ideal(x, conf.int=0.95, start=rownames(x$x)[1], ...)

Arguments

x

an object of class ideal.

conf.int

for "ideal" objects with 1 dimension estimated, the level of the confidence interval to plot around the posterior mean for each legislator. If 2 or more dimensions were estimated, conf.int is ignored.

start

the number of the iteration with which to begin when calculating the posterior mean and confidence interval.

...

other parameters to be passed through to plotting functions.

Details

If the ideal object comes from fitting a d=1 dimensional model, then plot.ideal() produces a plot of the posterior mean of the ideal point estimations for each legislator with a confidence interval. This is the same graph that would be produced by calling plot1d. If there are more than 30 legislators, only 30 points on this graph will be labelled. If the ideal object has d=2 dimensions, then plot2d is called, which plots the (estimated) mean of the posterior density of each legislator (i.e., the ideal point is a point in 2-dimensional Euclidean space, and the posterior density for each ideal point is a bivariate density).

For unidimensional and two-dimensional models, if party information is available in the rollcall object contained in the ideal object, legislators from different parties are plotted in different colors. If the ideal object has more than 2 dimensions, plot.ideal() produces a matrix of plots of the mean ideal points of each dimension against the posterior mean ideal points of the other dimensions.

See Also

ideal, plot1d, plot2d

Examples

data(s109)
id1 <- ideal(s109,
             d=1,
             meanzero=TRUE,
             store.item=TRUE,
             maxiter=1000,   ## short run for examples
             burnin=100,
             thin=10)  

plot(id1)