get.corr.key {BoutrosLab.plotting.general}

Correlation Key

Description

A function for adding correlation key legends to scatterplots.

Usage

get.corr.key(
	x, 
	y, 
	label.items = c("spearman", "spearman.p"), 
	x.pos = 0.03, 
	y.pos = 0.97, 
	key.corner = NULL, 
	key.cex = 1, 
	key.title = NULL, 
	title.cex = 1, 
	alpha.background = 0,
	num.decimals = 2,
        border = 'white'
	)

Arguments

x

A vector of values

y

Another vector of values with the same length as x

label.items

A vector of things to include in the key. Any combination of the following can be used. c("spearman","pearson","kendall","beta0", beta1,"spearman.p","pearson.p","kendall.p","beta1.p", "beta.robust", "beta.robust.p"). "all" is an alternative to the full list.

x.pos

Horizontal position of the key corner

y.pos

Vertical position of the key corner

key.corner

The corner of the key defaults to the closest corner of the plot. This helps overcome some variable character and row sizing.

key.cex

Specifies the size of font for the key, defaults to 1.

key.title

The title of the key. Defaults to NULL

title.cex

The size of the key title. Defaults to 1

alpha.background

A value from 0 to 1 indicating the transparency of the legend box.

num.decimals

Number of decimal places to keep for spearman, pearson and kendall correlations. Defaults to 2.

border

Adds border around the key with the color specified, alpha background cannot be 0. Defaults to White.

Value

Returns a key in the format specified in the xyplot documentation.

Author(s)

Daryl Waggott

See Also

xyplot, plotmath

Examples

# create some temporary data
tmp.data <- data.frame(
    x = c(
        runif(n = 15, min =  0, max =  20),
        runif(n = 15, min = 80, max = 100),
        runif(n = 70, min =  0, max = 100)
        ),
    y = c(
        runif(n = 15, min =  0, max =  20),
        runif(n = 15, min = 80, max = 100),
        runif(n = 70, min =  0, max = 100)
        )
    );

# a simple scatterplot with correlation key
BoutrosLab.plotting.general::create.scatterplot(
    formula = y ~ x,
    data = tmp.data,
    # filename = tempfile(pattern = 'get.corr.key-scatterplot', fileext = '.tiff'),
    xlab.label = 'X Axis Title',
    ylab.label = 'Y Axis Title',
    xlimits = c(0,100),
    ylimits = c(0,100),
    xat = seq(0,100,25),
    yat = seq(0,100,25),
    add.axes = FALSE,
    key = BoutrosLab.plotting.general::get.corr.key(
        tmp.data$y,
        tmp.data$x,
        label.items = c('spearman', 'spearman.p', 'kendall', 'beta1', 'beta1.p')
        )
    );


# compare beta1 vs a robust estimate of the slope

# add an outlier
tmp.data <- rbind(tmp.data, c(2000,100));

BoutrosLab.plotting.general::create.scatterplot(
    formula = y ~ x,
    data = tmp.data,
    # filename = tempfile(pattern = 'get.corr.key.robust-scatterplot', fileext = '.tiff'),
    xlab.label = 'X Axis Title',
    ylab.label = 'Y Axis Title',
    xlimits = c(0,100),
    ylimits = c(0,100),
    xat = seq(0,100,25),
    yat = seq(0,100,25),
    add.axes = FALSE,
    key = BoutrosLab.plotting.general::get.corr.key(
        tmp.data$y,
        tmp.data$x,
        label.items = c('beta1', 'beta1.robust','beta1.p','beta1.robust.p')
        )
    );

# see create.scatterplot for an example of creating multiple keys using legends