sciplot (version 1.2-0)

bargraph.CI: Bar Graph of Calculated Means and CI

Description

Barplot of the mean and standard error (or other summary statistics) of a response variable for one-way or higher experimental designs.

Usage

bargraph.CI(x.factor, response, group=NULL, split=FALSE,
    xlab=NULL, ylab=NULL, col=NULL, angle=NULL, density=NULL,
    lc=TRUE, uc=TRUE, legend=FALSE, ncol=1,
    leg.lab=NULL, x.leg=NULL, y.leg=NULL, cex.leg=1,
    bty="n", bg="white", space=if(split) c(-1,1),
    err.width=if(length(levels(as.factor(x.factor)))>10) 0 else .1,
    err.col="black", err.lty=1,
    fun = function(x) mean(x, na.rm=TRUE),
    ci.fun= function(x) c(fun(x)-se(x), fun(x)+se(x)),
    ylim=NULL, xpd=FALSE, data=NULL, subset=NULL, ...)

Arguments

x.factor

a factor (required) whose levels will form the x axis.

response

a numeric variable giving the response.

group

grouping factor (optional) whose levels will form groups of bars for each level of x.factor.

split

logical. Should groups of bars be displayed back to back. See details below.

xlab

default x-axis label. If not specified will parse name from x.factor. To leave blank set to NA

ylab

default y-axis label. If not specified will parse name from response. To leave blank set to NA

col

default color(s) for bars.

angle

default angle of shading lines.

density

default density of shading lines.

lc, uc

logical. Should upper or lower CI be drawn?

legend

logical. Should a legend be included?

ncol

number of columns to use for legend.

leg.lab

legend labels for trace factors.

x.leg, y.leg

optional values to over ride the default legend placement.

cex.leg

character expansion value for legend labels.

bty

line type of bounding box for legend. Defaults to 'none'.

space

set spacing for bars.

err.width

set width of whiskers for error bars.

err.col

color for error bars. Defaults to "black".

err.lty

line type for error bars.

bg

background color of legend.

fun

the function to compute the summary statistic. Should return a single real value. Defaults to mean with NA values removed.

ci.fun

the functions to compute the CI. Should return a vector of length 2 defining the lower and upper limits of the CI's. Defaults to the mean +/- 1 standard error, with NA values removed.

ylim

range of y axis.

xpd

defualts to xpd=FALSE (in contrast to barplot) to cut off the range of data plotted.

data

an optional data frame.

subset

an optional expression indicating the subset of the rows of 'data' that should be used in the plot.

...

further graphical parameters. See barplot() for additional options.

Details

Plots a response as a function of treatment (factor) combinations for one-way and higher designs. This is a "wrapper" function for barplot that adds confidence intervals - barplot should be consulted for details. Note that the option "beside=TRUE" from barplot is hardcoded and that the default for value xpd is FALSE. If the option split is specified, the response associated with the second grouping factor will be made negative and displayed back-to-back with the first.

See Also

barplot, se

Examples

# NOT RUN {
data(ToothGrowth)

# One way design
bargraph.CI(x.factor = dose, response = len, data = ToothGrowth) 

# Two-way design with options
bargraph.CI(dose, len, group = supp, data = ToothGrowth,
            xlab = "Dose", ylab = "Growth", cex.lab = 1.5, x.leg = 1,
            col = "black", angle = 45, cex.names = 1.25,
            density = c(0,20), legend = TRUE)
# }