plot.gsbMainOut: Plot methods
In gsbDesign: Group Sequential Bayes Design

Description Usage Arguments Value Author(s) References See Also Examples

View source: R/package.r

Methods for plotting the results of gsb().

## S3 method for class 'gsbMainOut'
plot(x,
     what=c("all", "cumulative all",
            "both", "cumulative both",
            "sample size", "success", "futility",
            "success or futility", "indeterminate", "cumulative success",
            "cumulative futility", "cumulative success or futility",
            "cumulative indeterminate", "boundary",
            "std.boundary","delta.grid","patients"),
     range.delta = "default",
     stages = "default",
     delta.grid = TRUE,
     color = TRUE,
     smooth = 100,
     contour = TRUE, 
     export = FALSE,
     path = tempdir(),
     sliced = FALSE,
     range.control="default", ...)

## S3 method for class 'gsbSimulation'
plot(x,...)

## S3 method for class 'gsbDesign'
plot(x,...)

`x`	object of appropriate class.
`what`	a `character string` to choose a plot. It should correspond to one level of `OC$type` where `OC` is an object of the output of `gsb()`. Additional possibilities are `what = "all"` to plot the success-, futility- and success or futility-probabilities, `what = "cumulative all"` to plot the cumulative success-, cumulative futility- and cumulative success or futility-probabilities, `what = "boundary"` or `what = "std.boundary"` to plot the bounds, `what = "patients"` for a histogram of the patients per stage and `what = "delta.grid"` to plot the grid of delta's.
`range.delta`	a `vector` of length 2. For choosing the plot limits manually set `range.delta = c(min, max)` for a Bayesian update on "treatment effect" or `range.delta = c(control.min, control.max, treatment.min, treatment.max)` for a Bayesian update "per arm". If `range.delta = "default"` the range of the plot is chosen so that all delta's are covered.
`stages`	a `vector` of length 2 containing the number of the lowest and highest stage, which should be plotted. If `stages = "default"` all stages are plotted (expect for `what = "sample size"` where only the last stage is plotted.)
`delta.grid`	`logical`. If `TRUE` the delta grid is plotted too.
`color`	`logical`. If `TRUE` the plot is colored.
`smooth`	a `vector` of length 1. A higher number makes the plot 'smoother' if `type.update = "per arm"`.
`contour`	`logical`. If `TRUE` contour lines are added to the plot.
`export`	`logical`. If `TRUE` the plot is save as .png-file.
`path`	`character`. to specify the location to which to table should be exported. the default 'tempdir()' exports the table to a temporary directory.
`sliced`	`logical`. If `TRUE` the contour plot for `type.update = "per arm"` is shown in several 2D plots. In order to use this option the argument `grid.type` of `gsbSimulation()` has to be `sliced`.
`range.control`	a `vector` of length 2 or \"default\". If `special = TRUE` the range of the control values can be set manually as vector c(min, max).
`...`	further arguments passed to or from other methods.

Returns an object of class "trellis"

Florian Gerber <florian.gerber@math.uzh.ch>, Thomas Gsponer

uses the R-package 'lattice'.

gsb, xyplot

## please see examples of function 'gsb'. 
## --------------------------------------


## --------------------------------------
## alternative plots can be created for example
## with package 'ggplot2'.

des <- gsbDesign(nr.stages=2,
                     patients=10,
                     sigma=10,
                     criteria.success=c(0,0.8, 7, 0.5),
                     criteria.futility=c(2,0.8),
                     prior.difference="non-informative")

sim <- gsbSimulation(truth=c(-10,20,60),
                             type.update="treatment effect")


x <- gsb(des,sim)

## get data.frame with operating characteristics
datgraph <- x$OC


## prepare for plot
sub <- c("success", "futility", "success or futility")
datgraph2 <- subset(datgraph,datgraph$type %in% sub)
datgraph2$type <- as.factor(paste(datgraph2$type))
datgraph2$value[datgraph2$type=="cumulative success or futility"] <-
1-datgraph2$value[datgraph2$type=="cumulative success or futility"]
levels(datgraph2$type) <- c("1)cumulative futility" ,"3)cumulative success","2)indeterminate")
datgraph2$type=as.factor(paste(datgraph2$type))
levels(datgraph2$type) <- c("cumulative futility" ,"indeterminate","cumulative success")
datgraph2 <- datgraph2[order(datgraph2$delta),]


## plots
p1 <- qplot(delta,value,geom="blank",color=type,facets=.~stage,data=datgraph2,
xlab=expression(delta))

p1+geom_line(size=1.5)+scale_color_manual(values = c("cumulative futility" = "dark red",
"indeterminate" = "orange", "cumulative success" = "dark green"))

p2=p1+geom_area(aes(x = delta,y=value,fill=type))

p2+scale_fill_manual(values = c("cumulative futility" = "dark red",
"indeterminate" = "orange", "cumulative success" = "dark green"))