plot.norMmix {norMmix}

Plot Method for "norMmix" Objects

Description

This is the S3 method for plotting "norMmix" objects.

Usage

## S3 method for class 'norMmix'
plot(x, y=NULL, ...) 
## S3 method for class 'norMmixMLE'
plot(x, y = NULL,
     show.x = TRUE,
     main = sprintf(
         "norMmixMLE(*, model=\"%s\") fit to n=%d observations in %d dim.",
         nm$model, x$nobs, nm$dim
     ),
     sub = paste0(
         sprintf("log likelihood: %g; npar=%d", x$logLik, x$npar), 
         if (!is.null(opt <- x$optr)) paste("; optim() counts:", named2char(opt$counts))
     ),
     cex.data = par("cex") / 4, pch.data = 4,
     ...)
 
## S3 method for class 'fittednorMmix'
plot(x, main = "unnamed", plotbest = FALSE, ...)

plot2d (nMm, data = NULL,
        add = FALSE,
        main = NULL,
        sub = NULL,
        type = "l", lty = 2, lwd = if (!is.null(data)) 2 else 1,
        xlim = NULL, ylim = NULL, f.lim = 0.05,
        npoints = 250, lab = FALSE,
        col = Trubetskoy10[1],
        col.data = adjustcolor(par("col"), 1/2),
        cex.data = par("cex"), pch.data = par("pch"),
        fill = TRUE, fillcolor = col, border = NA,
        ...)
plotnd(nMm, data = NULL,
       main = NULL,
       diag.panel = NULL,
       ...)
Trubetskoy10

Arguments

x, nMm

an R object inheriting from "norMmix".

y

further data matrix, first 2 columns will be plotted by "points"

...

further arguments to be passed to another plotting function.

show.x

Option for plot.norMmixMLE. Plot data points along with estimated model. Defaults to TRUE.

data

Data points to plot.

add

This argument is used in the internal function, plot2d, to control whether to create a new plot or add to an existing one. Should not be set by the user. Defaults to FALSE

main

Set main title. See Usage section for default values.

sub

Set subtitle. See Usage section for default values.

type

Graphing type for ellipses border. Defaults to "l".

lty

Line type to go with the type. See "par".

lwd

Line width as in lty.

xlim

Set explicit x limits for 2d plots.

ylim

As xlim.

f.lim

Percentage value for how much to extend xlim and ylim. As in the f argument to "extendrange".

npoints

How many points to use in the drawn ellipses. Larger values make them prettier but might affect plot times.

lab

Whether to print labels for mixture components. Will print "comp

col

Fill color for ellipses. Default is "#4363d8".

col.data

Color to be used for data points.

cex.data

See "par".

pch.data

See "par".

fill

Leave ellipses blank with outline or fill them in.

fillcolor

Color for infill of ellipses.

border

Argument to be passed to polygon.

diag.panel

Function to plot 2d projections of a higher-dimensional mixture model. Used by plotnd. Requires function with signature function(x, y, data = NULL, ...) Should not be set by the user.

plotbest

Used by fittednorMmix. Plot best fitting model using plot.norMmix.

Details

The plot method calls one of two auxiliary functions, one for dim=2, another for higher dimensions. The method for 2 dimensional plots also takes a add parameter (FALSE by default), which allows for the ellipses to be drawn over an existing plot.

The higher dimensional plot method relies on the pairs.default function to draw a lattice plot, where the panels are built using the 2 dimensional method.

Trubetskoy10: A vector of colors for these plots, chosen to be distinguishable and accessible for the colorblind, according to https://sashamaps.net/2017/01/11/list-of-20-simple-distinct-colors/, slightly rearranged, so that the first five colors stand out well on white background.

Value

plot.norMmix In the 2 dimensional case, returns invisibly coordinates of bounding ellipses of distribution.

Examples

plot(MW212) ## and add a finite sample realization:
points(rnorMmix(n=500, MW212))

## or:
x <- points(rnorMmix(n=500, MW212))
plot(MW212, x)


## Example of dim. = p > 2 :
plot(MW34)