car (version 2.1-6)

scatterplotMatrix: Scatterplot Matrices


Enhanced scatterplot matrices with univariate displays down the diagonal; spm is an abbreviation for scatterplotMatrix. This function just sets up a call to pairs with custom panel functions.


scatterplotMatrix(x, ...)

# S3 method for formula scatterplotMatrix(formula, data=NULL, subset, labels, ...)

# S3 method for default scatterplotMatrix(x, var.labels=colnames(x), diagonal=c("density", "boxplot", "histogram", "oned", "qqplot", "none"), adjust=1, nclass, plot.points=TRUE, smoother=loessLine, smoother.args=list(), smooth, span, spread = !by.groups, reg.line=lm, transform=FALSE, family=c("bcPower", "yjPower"), ellipse=FALSE, levels=c(.5, .95), robust=TRUE, groups=NULL, by.groups=FALSE, use=c("complete.obs", "pairwise.complete.obs"), labels, id.method="mahal", id.n=0, id.cex=1, id.col=palette()[1], id.location="lr", col=if (n.groups == 1) palette()[3:1] else rep(palette(), length=n.groups), pch=1:n.groups, lwd=1, lty=1, cex=par("cex"), cex.axis=par("cex.axis"), cex.labels=NULL, cex.main=par("cex.main"), legend.plot=length(levels(groups)) > 1, legend.pos=NULL, row1attop=TRUE, ...)

spm(x, ...)



a data matrix, numeric data frame.


a one-sided “model” formula, of the form ~ x1 + x2 + ... + xk or ~ x1 + x2 + ... + xk | z where z evaluates to a factor or other variable to divide the data into groups.


for scatterplotMatrix.formula, a data frame within which to evaluate the formula.


expression defining a subset of observations.


Arguments for the labelling of points. The default is id.n=0 for labeling no points. See showLabels for details of these arguments. If the plot uses different colors for groups, then the id.col argument is ignored and label colors are determined by the col argument.


variable labels (for the diagonal of the plot).


contents of the diagonal panels of the plot. If plotting by groups, a different univariate display (with the exception of "histogram") will be drawn for each group.


relative bandwidth for density estimate, passed to density function.


number of bins for histogram, passed to hist function.


if TRUE the points are plotted in each off-diagonal panel.


a function to draw a nonparametric-regression smooth; the default is gamLine, which uses the gam function in the mgcv package. For this and other smoothers, see ScatterplotSmoothers. Setting this argument to something other than a function, e.g., FALSE suppresses the smoother.


a list of named values to be passed to the smoother function; the specified elements of the list depend upon the smoother (see ScatterplotSmoothers).

smooth, span

these arguments are included for backwards compatility: if smooth=TRUE then smoother is set to loessLine, and if span is specified, it is added to smoother.args.


if TRUE, estimate the (square root) of the variance function. For loessLine and for gamLine, this is done by separately smoothing the squares of the postive and negative residuals from the mean fit, and then adding the square root of the fitted values to the mean fit. For quantregLine, fit the .25 and .75 quantiles with a quantile regression additive model. The default is TRUE if by.groups=FALSE and FALSE is by.groups=TRUE.


if not FALSE a line is plotted using the function given by this argument; e.g., using rlm in package MASS plots a robust-regression line.


if TRUE, multivariate normalizing power transformations are computed with powerTransform, rounding the estimated powers to `nice' values for plotting; if a vector of powers, one for each variable, these are applied prior to plotting. If there are groups and by.groups is TRUE, then the transformations are estimated conditional on the groups factor.


family of transformations to estimate: "bcPower" for the Box-Cox family or "yjPower" for the Yeo-Johnson family (see powerTransform).


if TRUE data-concentration ellipses are plotted in the off-diagonal panels.


levels or levels at which concentration ellipses are plotted; the default is c(.5, .9).


if TRUE use the cov.trob function in the MASS package to calculate the center and covariance matrix for the data ellipses.


a factor or other variable dividing the data into groups; groups are plotted with different colors and plotting characters.


if TRUE, regression lines are fit by groups.


if "complete.obs" (the default), cases with missing data are omitted; if "pairwise.complete.obs"), all valid cases are used in each panel of the plot.


plotting characters for points; default is the plotting characters in order (see par).


colors for lines and points; the default is taken from the color palette, with palette()[3] for linear regression lines, palette()[2] for nonparametric regression lines, and palette()[1] for points if there are no groups, and successive colors for the groups if there are groups.


width of linear-regression lines (default 1).


type of linear-regression lines (default 1, solid line).

cex, cex.axis, cex.labels, cex.main

set sizes of various graphical elements (see par).


if TRUE then a legend for the groups is plotted in the first diagonal cell.


position for the legend, specified as one of the keywords accepted by legend. If NULL, the default, the position will vary by the diagonal argument --- e.g., "topright" for diagonal="density".


If TRUE (the default) the first row is at the top, as in a matrix, as opposed to at the bottom, as in graph (argument suggested by Richard Heiberger).


arguments to pass down.


NULL. This function is used for its side effect: producing a plot.


Fox, J. and Weisberg, S. (2011) An R Companion to Applied Regression, Second Edition, Sage.

See Also

pairs, scatterplot, dataEllipse, powerTransform, bcPower, yjPower, cov.trob, showLabels, ScatterplotSmoothers.


Run this code
scatterplotMatrix(~ income + education + prestige | type, data=Duncan)
scatterplotMatrix(~ income + education + prestige, 
    transform=TRUE, data=Duncan, smoother=loessLine)
scatterplotMatrix(~ income + education + prestige | type, smoother=FALSE,, transform=TRUE, data=Duncan)
# }

Run the code above in your browser using DataCamp Workspace