plot.mleBb {dbd}

Plot a maxium likelihood fit to data from a beta binomial distribution.

Description

Creates a plot of type "h" of the probabilities of each possible x value of a beta binomial distribution where the probabilities are calculated on the basis of parameters estimated by the function mleBb(). If obsd is supplied it also superimposes/juxtaposes vertical lines representing the observed proportions.

Usage

## S3 method for class 'mleBb'
plot(x, ..., plot = TRUE, col.fit = "red", col.obsd = "blue",
              tikx = NULL, xlim=NULL, ylim=NULL, xlab = NULL,
              ylab = NULL, obsd = NULL, incr = NULL,
              main = "", legPos = "topright")

Arguments

x

An object of class "mleBb" as returned by the function mleBb()

...

Not used.

plot

Logical scalar; should a plot be produced (or should the function simply return a data frame consisting of the relevant values)?

col.fit

The colour for the (vertical) lines corresponding to the “fitted” probabilities, i.e. the probabilities calculated from the fitted parameters.

col.obsd

The colour for the (vertical) lines corresponding to the “observed” probabilities (proportions), i.e. the probabilities calculated by tabulating the data (from which the parameters were estimated. Ignored if obsd is not supplied.

tikx

(Optional) vector of locations of the tick marks on the x-axis.

xlim

A numeric vector of length 2 specifying the limits of the x-axis. Defaults to c(0,size) Note that \ codesize is extracted from argument x.

ylim

A numeric vector of length 2 specifying the limits of the y-axis. There is a “sensible” default.

xlab

A label for the x-axis; defaults to x.

ylab

A label for the y-axis; defaults to probability.

obsd

The data set from which the parameters were estimated, i.e. from which x was obtained. (Optional.)

incr

Numeric scalar; defaults to 0.1 if size (extracted from x) is less than 20 and to 0.5 otherwise. This number should be non-zero and less than 1 in absolute value. (One would usually want it to be positive, but it could conceiveably be set to a small negative value.) It gives the value of the “increment” or “shift” that separates the vertical lines representing the fitted probabilities and those representing the observed proportions (calculated from obsd). Ignored if obsd is not supplied.

main

A main title for the plot; defaults to the empty string.

legPos

A list with components x and y, or a text string, specifying the placement of the legend. See legend() for details. A legend is plotted only if obsd is specified, whence legPos is otherwise ignored. The plotting of a legend may be suppressed (even when obsd is supplied) by setting legPos=NULL.

Value

A data frame with numeric columns x, p and possibly po. The x column consists of the integers from 0 to size. The p column consists of the appropriate probabilities of the x values, calculated by dbetabinom() from the rmutil package. The po column is present only if obsd is supplied and consists of the observed proportions. The value is returned invisibly. A plot is produced as a side-effect if plot is TRUE.

Note

This function calls plotBb() to do the heavy lifting.

Warning

It is up to the user to make sure that the obsd argument, if specified, is indeed the data set from which the object x was calculated.

Author(s)

Rolf Turner [e-mail]

See Also

mleBb() plotBb()

Examples

   if(require(hmm.discnp)) {
      xxx <- hmm.discnp::Downloads
      fit <- mleBb(xxx,size=14)
      plot(fit)
      plot(fit,obsd=xxx)
      plot(fit,obsd=xxx,legPos=list(x=3,y=0.25))
      plot(fit,obsd=xxx,legPos=NULL) # No legend is plotted.
   }
   set.seed(42)
   yyy <- rbinom(300,10,0.7)
   fit <- mleBb(yyy,size=10)
   plot(fit,obsd=yyy,legPos="topleft")