ggplot.default {ggplot} | R Documentation |
Create a new ggplot plot
ggplot.default(data = NULL, formula = . ~ ., margins=FALSE, aesthetics=list(), ...)
data |
default data frame |
formula |
formula describing row and column layout, see reshape for more details |
margins |
a vector of names giving which margins to display, can include grand_row and grand_col or uss TRUE to display all margins |
aesthetics |
default list of aesthetic mappings (these can be colour, size, shape, line type – see individual grob functions for more details) |
... |
This function creates the basic ggplot object which you can then
furnish with graphical objects. Here you will set
up the default data frame, default aesthetics and the formula that
will determine how the panels are broken apart. See reshape
for more details on specifying the facetting formula and margin arguments.
Note that ggplot creates a plot object without a "plot": you need to
grobs (points, lines, bars, etc.) to create something that you can see.
To get started, read the introductory vignette: vignette("introduction", "ggplot")
Steps to create a plot:
p <- ggplot(mtcars, aesthetics=list(y=hp, x=mpg))
)
ggpoint(p)
)
or, use qplot
Simple grobs:
ggabline
: line with given slope and intercept
ggarea
: area (polygons with base on y=0)
ggbar
: bars (stocked and dodgted)
ggjitter
: jittered points (useful for discrete data)
ggline
: lines (paths sorted by x-axis values)
ggpath
: paths
ggpoint
: points
ggribbon
: ribbon
ggtext
: text
ggtile
: tiles, like a levelplot
Complex grobs:
ggboxplot
: box plot
ggcontour
: contour lines
ggdensity
: 1d density plot (continuous analogue of histogram)
gg2density
: 2d density countours
gghistogram
: histogram
ggquantile
: quantile lines from a quantile regression
ggsmooth
: smooths from any model family
Look at the documentation of these objects to see many examples of ggplot in action.
You will also want to add scales to the basic plot to give finer control over how the data values are mapped to aethetics attributes of the grobs. For scales that control position of the points see:
pscontinuous
: continuous scales (with optional transformation)
pscategorical
: categorical scales
For other scales, see:
sccolour
: colour categorical variables using Brewer colour scales (see also scfill
)
scgradient
: colour continuous scales with a gradient (see also scfillgradient
)
schcl
: map continuous variable to hue, chroma or luminance components (see also scfillhcl
)
schsv
: map continuous variable to hue, saturation or value components (see also scfillhsv
)
scmanual
: no automatic conversion, uses raw values directly
sclinetype
: line type (solid, dashed, dotted, etc.)
scrgb
: map continuous variable to red, green or blue components (see also scfillrgb
)
scshape
: point shape (glyph)
scsize
: point or line size
ggplot is different from base and lattice graphics in how you build up the plot. With ggplot you build up the plot object (rather than the plot on the screen as in base graphics, or all at once as in lattice graphics.)
Each of the grob and scale functions adds the grob to the plot and returns the modified plot object. This lets you quickly experiment with different versions of the plot, using different grobs or scales. You can see how this works in the examples
You can also use summary
to give a quick description of a plot.
If you want to change the background colour, how the panel strips are displayed,
or any other default graphical option, see ggopt
.
Hadley Wickham <h.wickham@gmail.com>
http://had.co.nz/ggplot, stamp
, reshape
, ggopt
, vignette("introduction", "ggplot")
p <- ggplot(tips) summary(p) ggpoint(p, aesthetic=list(y = tip, x=total_bill)) p <- ggplot(tips, aesthetic=list(y = tip, x=total_bill)) p$title <- "Tips" summary(p) ggpoint(p) ggpoint(p, colour="darkgreen", size=3) ggpoint(p, list(colour=sex)) ggpoint(ggplot(tips, . ~ sex,aesthetics = list(y = tip, x = total_bill))) p <- ggplot(tips, smoker ~ sex,aesthetics = list(y = tip, x = total_bill)) ggpoint(p) ggsmooth(ggpoint(p)) ggsmooth(ggpoint(p), method=lm, formula=y~x) ggabline(ggpoint(p), slope=c(0.1,0.15,0.2)) (p2 <- ggabline(ggpoint(p, aes=list(colour=tip/total_bill)), slope=c(0.1,0.15,0.2))) summary(p2) scgradient(p2) scgradient(p2, midpoint=0.15, high="green", mid="yellow") p<-ggplot(tips, sex ~ smoker, aesthetics=list(x=tip/total_bill), margins=TRUE) gghistogram(p) gghistogram(p,scale="density", breaks=seq(0,1, length=20)) ggdensity(gghistogram(p)) p<-ggplot(tips, . ~ smoker, aesthetics=list(x=sex, y=tip)) ggboxplot(p) ggjitter(ggboxplot(p))