Draws a bubble plot where the position is given by x and y,
and the size of each bubble increases with area.
Usage
plotBubble(x, ...)
# Default S3 method
plotBubble(
x,
y,
area,
main = "",
xlab = "",
ylab = "",
xlim = NULL,
ylim = NULL,
add = FALSE,
col = NA,
border = NULL,
cex = 1,
grid = NA,
na.rm = FALSE,
stamp = .useTheme,
...
)
# S3 method for class 'formula'
plotBubble(
formula,
data = NULL,
subset,
na.action = na.omit,
main = "",
xlab = "",
ylab = "",
xlim = NULL,
ylim = NULL,
add = FALSE,
col = NA,
border = NULL,
cex = 1,
grid = NA,
...
)Arguments
- x
numeric vector of x positions.
- ...
additional graphical parameters passed to
par().- y
numeric vector of y positions.
- area
non-negative numeric vector controlling the bubble sizes.
- main, xlab, ylab
plot labels.
- xlim, ylim
axis limits.
- add
logical; if
TRUE, adds to an existing plot.- col
fill color(s) of the bubbles.
- border
border color(s) of the bubbles.
- cex
scaling factor applied to the bubble radii.
- grid
logical,
NA, or list controlling background grid.- na.rm
logical. Observations with a missing
x,yorareacannot be drawn and are always omitted; withna.rm = FALSE(default) a warning reports how many. Missing colors are not missing data,NAdraws transparent.- stamp
controls the corner stamp.
.useTheme(default) resolves togetTheme()$stamp.TRUE/FALSE/NULL, a string, or a named list forstamp().- formula
A formula of the form
y ~ x | area.- data
optional data frame.
- subset
optional subset expression, evaluated in
data.- na.action
function to handle missing values, defaults to
na.omit().
Value
Invisibly, a data frame with one row per drawn bubble and the
columns x, y, area, col, border, and the radii rx, ry in
user coordinates, sorted by decreasing area.
Details
The function supports both a default interface and a formula interface
of the form y ~ x | area.
The bubble radii are scaled relative to the largest area: the largest
bubble gets a radius of 0.35 * cex inches, the others
\(r = 0.35 \cdot cex \cdot (area / \max(area))^{0.4}\). The exponent
0.4 (instead of 0.5 for strictly area-proportional circles) dampens the
size differences, so that small bubbles remain visible. Radii are
converted to user coordinates separately for both axes, so the bubbles
are drawn as circles regardless of the aspect ratio.
Bubbles are drawn in order of decreasing area, so small bubbles are not hidden behind large ones.
Graphical elements such as grids are controlled via the unified plot
design system using bedrock::callIf() and .theme().
See also
Other plot.bivariate:
plotAssoc(),
plotBag(),
plotCor(),
plotDens2D(),
plotHeatmap(),
plotHexbin(),
plotMosaic(),
plotPyramid(),
plotXY()
Examples
set.seed(1)
x <- rnorm(100)
y <- rnorm(100)
a <- runif(100, 1, 10)
plotBubble(x, y, a)
df <- data.frame(x = x, y = y, a = a)
plotBubble(y ~ x | a, data = df)
US <- data.frame(state.x77, State=state.name,
Region=state.region, Abb=state.abb)
plotBubble(Income ~ Population | Area ,
data=US,
grid=TRUE, col=addOpacity(pal("helsana")[US$Region]), cex=1.2 )
text(Income ~ Population, US, labels=US$Abb, cex=0.8)
