Points: Number of cases in each location reflected in the point size

Description

Number of cases in each location reflected in the point size

Usage

Points(x, y, pch=1, centers=FALSE, scale=1, cex.min=1, col=1,
 na.omit=TRUE, plot=TRUE, ...)
PPoints(groups, x, y, cols=as.numeric(groups), pchs=as.numeric(groups),
 na.omit.all=TRUE, ...)

Arguments

x, y

Coordinates

pch

Point type

pchs

Types of point groups

centers

If TRUE, show centers of each location as a pixel-size dot (pch=".")

cex.min

Minimal point size

col

Color of points

cols

Color of point groups

na.omit

If TRUE (default), skip data points with NAs

plot

If FALSE, does not plot

na.omit.all

If TRUE (default), skip data points and corresponding factor values with NAs, then make 'na.omit' for internal Points() FALSE

scale

Scale factor for point size

groups

Factor defining groups

...

Points() passes other arguments to points(), PPoints() passes other arguments to Points()

Value

Invisibly returns vector of "multiplication indexes", in case of PPoints() it is group-wise so overplotting between groups does not count. Please keep in mind that these indexes only indicate how many times the point is overplotted, but do not show groups of duplicates. Use Alldups() for groups.

Details

Frequently, more then one data point is located in one coordinate place (so called "overplotting"). How to show overplotting? One way is 'jitter()', these is also (really advanced) 'sunflowerplot()'. 'Points()' does it in its own way: number of cases in each point will be reflected in the point size. 'Points()' is a low-level graphic function, analogous to 'points()'.

'PPoints()' is the same as 'Points()' but for multiple subgroups.

To prettify plot, it is recommended to change 'scale' and optionally also 'cex.min'.

Alternative is the base R 'sunflowerplot()' but it is hard to read and there is no possibility to show multiple groups in data. Another alternative might be points with transparent color.

Examples

Run this code

# NOT RUN {
## colors modified via palette()
plot(iris[, 1:2], type="n")
palette(rainbow(3))
PPoints(iris[, 5], iris[, 1], iris[, 2], pchs=0, scale=0.7)
palette("default")
## now with centers, colors default, pch by group, and one NA
iris[1, 1] <- NA
plot(iris[, 1:2], type="n")
PPoints(iris[, 5], iris[, 1], iris[, 2], scale=0.7, centers=TRUE)
data(iris) ## to restore default embedded object
# }