plot-methods {binseqtest} | R Documentation |
Methods for Function plot
and points
in Package binseqtest
Description
Plot binary sequential boundaries for "boundEst"
objects.
Usage
## S4 method for signature 'boundEst,missing'
plot(x,
rcol = c(orange = "#E69F00", blue = "#56B4E9", green = "#009E73"),
rpch = c(openCircle=1, filledCircle=16, filledDiamond=18),
bplottype = "NS",
newplot = TRUE, dtext=NULL, grid=50, xlab=NULL, ylab=NULL, ...)
## S4 method for signature 'boundEst'
points(x, ...)
Arguments
x |
an object of class |
rcol |
rejection color vector, rcol[1]=fail to reject, rcol[2]=reject, conclude theta>theta0, rcol[3]=reject, conclude theta< theta0 (see details) |
rpch |
rejection pch vector, correspond to same categories as rcol vector |
bplottype |
character, either 'NS' (default), 'FS', 'NB', 'NZ', or 'NE' (see details) |
newplot |
logical, should a new plot be started? if FALSE add to existing plot (only makes sense to add to plot with the same bplottype) |
dtext |
logical, add descriptive text? if NULL only adds text when newplot=TRUE (used for bplottype='NS' or 'FS') |
grid |
numeric, if maximum possible total trials<=grid then add gridlines (used for bplottype='NS' or 'FS') |
xlab |
title for x axis, if NULL value depends on bplottype |
ylab |
title for y axis, if NULL value depends on bplottype |
... |
other arguments to the |
Details
The default rcol vector are good colors for distinguishing for those with color blindness. Text is printed on the unused portion of the plot, which uses the color names taken from the rcol vector names.
Their are several different types of plots, selected by the bplottype
argument, where the value is a character string with 2 characters, the first representing the
x-axis and the second representing the y-axis. For example bplottype
='NS' denotes N=total number of trials on the horizontal axis, and S=number of successes on the vertical
axis. Other plots are: 'FS'=failure by successes; 'NB'=total by B-values; 'NZ'=total by Z-scores; 'NE'=total by estimates and confidence intervals. The type 'NE' is only defined
if there are only 1 value for each N on the upper and 1 value for each N on the lower part of the boundary. Otherwise, the confidence intervals would overlap and be uninformative.
For 'NE' the end of the boundary is not plotted because of that overlapping.
For some examples, see plot section of the vignette. The method points just calls plot(x,newPlot=FALSE,...)
.
Methods
signature(x = "ANY", y = "ANY")
Generic function: see
plot
.signature(x = "boundEst", y = "missing")
Plot binary sequential boundaries for
x
.signature(x = "ANY")
Generic function: see
points
.signature(x = "boundEst")
Add points associated with the binary sequential boundaries for
x
to a plot.
Examples
b<-designOBF(50,theta0=.5)
plot(b,bplottype="NE")
plot(b)
b2<-designFixed(49,theta0=.5)
points(b2,rpch=c(17,17,17))