Abbreviation:
Violin Plot only: vp, ViolinPlot
Box Plot only: bx, BoxPlot
Scatter Plot only: sp, ScatterPlot
A scatterplot displays the values of a distribution, or the relationship between the two distributions in terms of their joint values, as a set of points in an n-dimensional coordinate system, in which the coordinates of each point are the values of n variables for a single observation (row of data). From the identical syntax, from any combination of continuous or categorical variables variables x and y, Plot(x) or Plot(x,y), where x or y can be a vector, by default generates a family of related 1- or 2-variable scatterplots, possibly enhanced, as well as related statistical analyses. A categorical variable is either non-numeric, such as an R factor, or may be defined to consist of a small number of equally spaced integer values. The maximum number of such values to define such an integer variable as categorical is set by the n.cat parameter, with a default value of 0, that is, by default, all variables with numerical values are defined as continuous variables.
Plot is a general function, which produces a wide variety of scatterplots, which, for a single variable, can be in the context of violin plots and box plots, as outlined in the following list. The parameter definitions that follow this list are grouped, with parameters that relate to the same type of plot defined in the same group.
Plot(x,y): x and y continuous yields traditional scatterplot of two continuous variables
Plot(x,y): x and y categorical, to solve the over-plot problem, yields a bubble (balloon) scatterplot, the size of each bubble based on the corresponding joint frequency as a replacement for the two dimensional bar chart
Plot(x,y): x (or y) categorical and the other variable continuous, yields a scatterplot with means at each level of the categorical variable
Plot(x,y): x (or y) categorical with unique (ID) values and the other variable continuous, yields a Cleveland dot plot
Plot(X,y) or Plot(x,Y): one vector variable defined by several continuous variables, paired with another single continuous variable, yields multiple scatterplots on the same graph
Plot(x): one continuous variable generates either, a violin/box/scatterplot (VBS plot), introduced here, or a run chart with run=TRUE, or x can be an R time series variable for a time series chart
Plot(x): one categorical variable yields a 1-dimensional bubble plot to solve the over-plot problem for a more compact replacement of the traditional bar chart
Plot(X): one vector of continuous variables, with no y-variable, results in a scatterplot matrix
Plot(X): one vector of categorical x-variables, with no y-variable, generalizes to a matrix of 1-dimensional bubble plots, here called the bubble plot frequency matrix, to replace a series of bar charts
Represent the influence of additional categorical variables with by1 or by2 to generate Trellis plots conditioned on one or two variables from implicit calls to functions from Deepayan Sarkar's (2009) lattice package. Use by to group multiple variables on the same plot, or on multiple panels if Trellis graphics are activated. For a third variable, which is continuous, specify size for a bubble plot. By default, the values of analysis that generate the plotted points is data, or choose other values to plot, which are statistics computed from the data such as mean.
Plot(x, y=NULL, data=mydata, rows=NULL,
values=c("data", "count", "prop", "sum", "mean", "sd",
"min", "median", "max"),
n.cat=getOption("n.cat"), by=NULL, by1=NULL, by2=NULL,
n.row=NULL, n.col=NULL, aspect="fill",
fill=getOption("pt.fill"), color=getOption("pt.color"),
trans=getOption("trans.pt.fill"),
size=NULL, size.cut=NULL, shape="circle", means=TRUE,
sort.yx=FALSE, segments.y=FALSE, segments.x=FALSE,
jitter.x=0, jitter.y=0,
ID="row.name", ID.size=0.75,
MD.cut=0, out.cut=0, out.shape="circle", out.size=1,
vbs.plot="vbs", vbs.pt.fill=c("black", "default"), bw=NULL,
vbs.size=0.9, vbs.mean=FALSE, fences=FALSE,
k=1.5, box.adj=FALSE, a=-4, b=3,
radius=0.25, power=0.6,
low.fill=NULL, hi.fill=NULL, proportion=FALSE,
smooth=FALSE, smooth.points=100, smooth.trans=0.20,
smooth.bins=128,
fit="off", fit.se=0.95, ellipse=0,
bin=FALSE, bin.start=NULL, bin.width=NULL, bin.end=NULL,
breaks="Sturges", cumul=FALSE,
run=FALSE, lwd=2, area=FALSE, area.origin=0,
center.line=c("default", "mean", "median", "zero", "off"),
show.runs=FALSE, stack=FALSE,
xlab=NULL, ylab=NULL, main=NULL, sub=NULL,
lab.adj=c(0,0), margin.adj=c(0,0,0,0),
rotate.x=getOption("rotate.x"), rotate.y=getOption("rotate.y"),
offset=getOption("offset"),
xy.ticks=TRUE, value.labels=NULL, label.max=20, origin.x=NULL,
add=NULL, x1=NULL, y1=NULL, x2=NULL, y2=NULL,
auto=FALSE, eval.df=NULL, digits.d=NULL, quiet=getOption("quiet"),
do.plot=TRUE, width=NULL, height=NULL, pdf.file=NULL,
fun.call=NULL, …)
ScatterPlot(…)
sp(…)
BoxPlot(…)
bx(…)
ViolinPlot(…)
vp(…)
By itself, or with y, by default, a primary variable,
that is, plotted by its values mapped to coordinates.
The data values can be
continuous or categorical, cross-sectional or a time series.
If x is sorted, with equal intervals
separating the values, or is a time series, then by default
plots the points sequentially, joined by line segments.
Can specify multiple x-variables or multiple y-variables
as vectors, but not both. Can be in a data frame or defined
in the global environment.
An optional second primary variable. Variable with values to be mapped to coordinates of points in the plot on the vertical axis. Can be continuous or categorical. Can be in a data frame or defined in the global environment.
Optional data frame that contains one or both of x and
y. Default data frame is mydata.
A logical expression that specifies a subset of rows of the data frame to analyze.
The values that are the coordinates from which to plot the points,
data values by default. For y, which is continuous, then
for either a categorical x variable, or a continuous x variable
with values binned into categories, then can apply "mean", etc.
Number of categories, specifies the largest number of unique, equally spaced integer values of a variable for which the variable will be analyzed as categorical instead of continuous. Default is 0. Use to specify that such variables are to be analyzed as categorical, a kind of informal R factor.
A categorical variable to provide a scatterplot for
each level of the numeric primary variables x and
y on the same plot, a grouping variable.
For two-variable plots, applies to the panels of a
Trellis graphic if by1 is specified.
A categorical variable called a conditioning variable that
activates Trellis graphics, provided by Deepayan Sarkar's (2007) lattice
package, to provide
a separate scatterplot (panel) of numeric primary variables x
and y for each level of the variable.
A second conditioning variable to generate Trellis
plots jointly conditioned on both the by1 and by2 variables,
with by2 as the row variable, which yields a scatterplot (panel)
for each cross-classification of the levels of numeric x and
y variables.
Optional specification for the number of rows in the layout
of a multi-panel display with Trellis graphics. Specify
n.col or n.row, but not both.
Optional specification for the number of columns in the
layout of a multi-panel display with
Trellis graphics. Specify n.col or n.row, but
not both. If set to 1, then
the strip that labels each group locates to the left of each plot
instead of the top.
Lattice parameter for the aspect ratio of the panels,
defined as height divided by width.
The default value is "fill" to have the panels
expand to occupy as much space as possible. Set to 1 for square panels.
Set to "xy" to specify a ratio calculated
to "bank" to 45 degrees, that is, with the line slope approximately
45 degrees.
Transparency factor of the area of each point.
Default is
trans.pt.fill from the lessR style function.
When set to a constant, the scaling factor for standard points
(not bubbles) or a line, with default of 1.0 for points and 2.0 for a line.
Set to 0 to not plot the points or lines. When set to a variable, activates a
bubble plot with the size of each bubble further determined
by the value of radius. Applies to the standard two-variable
scatterplot as well as to the scatterplot component of the
integrated Violin-Box-Scatterplot (VBS) of a single continuous variable.
If TRUE (or 1), then for a bubble plot in which the
bubble sizes are defined by a
size variable, show the value
of the sizing variable for selected bubbles in the center of
the bubbles, unless the bubble is too small.
If FALSE, no value is displayed.
If a number greater than 1, then display the value only for the
indicated number of values, such as just the max and min for a setting of 2,
the default value when bubbles represent a size
variable. Color of the displayed text set by bubble.text from
the style function.
The plot character(s). The default value is a circle with both an
color and filled area, specified with color and fill.
Possible values are circle, square, diamond,
triup (triangle up), tridown (triangle down), all
uppercase and lowercase letters, all digits, and most punctuation characters.
The numbers 21 through 25 as defined by the R points function
also apply. If plotting levels according to by, then list one shape for
each level to be plotted.
If the one variable is categorical the other variable continuous,
then if TRUE, by default, plot means with the scatterplot. Also
applies to a 1-D scatterplot.
Sort the values of y by the values of x, such as
for a Cleveland dot plot, that is, a numeric x-variable paired
with a categorical y-variable with unique values. If a x
is a vector of two variables, sort by their difference.
For one x-variable, draw a line segment from the
y-axis to
each plotted point, such as for the Cleveland dot plot. For two
x-variables, the line segments connect the two points.
Draw a line segment from the x-axis for each plotted point.
Randomly perturbs the plotted points of a scatterplot horizontally according to an internally computed formula, or can be explicitly specified.
Randomly perturbs the plotted points of a scatterplot vertically according to an internally computed formula, or can be explicitly specified.
Name of variable to provide the labels for the plotted points, row names of data table (frame) by default.
Size of the plotted labels, with a default of 0.75 according
to the R parameter cex.
Modify text color of the labels with the style function
parameter ID.color.
Mahalanobis distance cutoff to define an outlier in a 2-variable scatterplot.
Count or proportion of plotted points to label, in order of their distance from the scatterplot center (means), counting down from the more extreme point. For two-variable plots, assess distance from the center with Mahalanobis distance. For VBS plots of a single continuous variable, refers to outliers on each side of the plot.
Shape of outlier points in a 2-variable scatterplot
or a VBS plot.
Modify fill color from the current theme with the
style function parameters out.fill and
out2.fill.
Size of outlier points in a 2-variable scatterplot or VBS plot.
A character string that specifies the components of the
integrated Violin-Box-Scatterplot (VBS) of a continuous variable.
A "v" in the string indicates a violin plot, a "b"
indicates a box plot with flagged outliers, and a "s"
indicates a 1-variable scatterplot. Default value is "vbs".
The characters can be in any order and upper- or lower-case.
Generalize to Trellis plots with the
by1 and by2 parameters, but currently only applies
to horizontal displays.
Modify fill and border colors from the current theme with
the style function parameters violin.fill,
violin.color, box.fill and box.color.
Points in a VBS scatterplot are black by default because
the background is the violin, which is based on the current theme
color. To use the values for pt.fill and pt.color
specified by the style function, set to "default".
Bandwidth for the smoothness of the violin plot. Higher values for smoother plots. Default is to calculate a bandwidth that provides a relative smooth density plot.
Width of the violin plot to the plot area. Make the violin (and also the accompanying box plot) larger or smaller by making the plot area and/or this value larger or smaller.
Show the mean on the box plot with a strip the color
of out.fill, which can be changed with the
style function.
If TRUE, draw the inner upper and lower fences as
dotted line segments.
IQR multiplier for the basis of calculating the distance of the whiskers of the box plot from the box. Default is Tukey's setting of 1.5.
Adjust the box and whiskers, and thus outlier detection, for skewness using the medcouple statistic as the robust measure of skewness according to Hubert and Vandervieren (2008).
Scaling factors for the adjusted box plot to set the length
of the whiskers. If explicitly set, activates box.adj.
Scaling factor of the bubbles in a bubble plot, which
sets the radius of the largest displayed bubble in inches, with default of
0.25 inches. To activate, set the value of size to
a third variable, which
sets the size of a bubble according to the size of the third variable.
Or activate when the values of the variables are categorical, either a
factor or an integer variable with the number of unique values less than
n.cat, in which case the size of the bubbles represents
frequency.
Relative size of the scaling of the bubbles to each other. Value of 0.5 scales the bubbles so that the area of each bubble is the value of the corresponding sizing variable. Value of 1 scales so the radius of the bubble is the value of the sizing variable, increasing the discrepancy of size between the variables. The default value is 0.6.
For a categorical variable and the resulting bubble plot, or a matrix of these plots, sets a color gradient of the fill color beginning with this color.
For a categorical variable and the resulting bubble plot, or a matrix of these plots, sets a color gradient of the fill color ending with this color.
Specify proportions, relative frequencies, instead of counts.
For a two variable bubble chart, if TRUE then to facilitate group
comparisons, displays the proportion of data values by fill variable
within each group.
Smoothed density plot for two numerical variables. By
default, set to TRUE for 2500 or more rows of data.
Number of points superimposed on the density plot in the areas of the lowest density to help identify outliers, which controls how dark are the smoothed points.
Exponent of the function that maps the density scale to the color scale.
Number of bins in both directions for the density estimation.
The best fit line. Default value is "off", with
options for non-linear"loess" and for linear least squares,
indicated by "lm". If
potential outliers are identified according to out.cut,
a second (dashed) fit line is displayed calculated without
the outliers. Modify the color from with the
style function parameter fit.color.
Confidence level for the error band displayed around the line of best fit. On by default at 0.95 if a fit line is specified. Can be a vector to display multiple ranges. Set to 0 to turn off.
Confidence level of a data ellipse for a scatterplot
of only a single
x-variable and a single y-variable according to the
contours of the corresponding bivariate normal density function. Can
specify the confidence level(s) for a single or vector of
numeric values from 0 to 1,
to plot one or more specified ellipses. For Trellis graphics, only the
maximum level applies with only one ellipse per panel.
Modify fill and border colors with the style function
parameters ellipse.fill and ellipse.color.
If TRUE, display the default frequency distribution
for the text output of the Violin-Box-Scatter (VBS) Plot, or,
if values is set to "count", a frequency polygon.
Optional specified starting value of the bins for a
frequency polygon or for the text output of a
Violin-Box-Scatter (VBS) Plot. Also, sets bin to TRUE.
Optional specified bin width value. Also, sets
bin to TRUE.
Optional specified value that is within the last bin, so the actual endpoint of the last bin may be larger than the specified value.
Specify a cumulative frequency polygon.
If set to TRUE, generate a run chart, i.e., line chart,
in which
points are plotted in the sequential order of occurrence in the data table.
By default, the points are connected by line
segments to form a run chart. Set by default when the x-values
are sorted with equal intervals or a single variable is a time series.
Customize the color of the line segments with segments.color
with function style.
Width of the line segments. Set to zero to remove the line segments.
If FALSE, no color is filled for the area under a plotted
line from a run chart or time series. Usual default is FALSE,
but default is TRUE if multiple time series are plotted on
the same panel. Default fill color is area.fill set with function style.
Origin for the filled area under the time series line. Values less than this value are below the corresponding reference line, values larger are above the line.
Plots a dashed line through the middle of a run chart. The two
possible values for the line are "mean" and "median".
Provides a center line for the "median" by default, when the values
randomly vary about the mean. A value of "zero" specifies the center
line should go through zero. Currently does not apply to Trellis plots.
If TRUE, display the individual runs in the run analysis.
Also, sets run to TRUE.
If TRUE, multiple time plots are stacked on each other, with
area set to TRUE by default.
Axis label for x-axis or y-axis.
If not specified, then the label becomes
the name of the corresponding variable label if it exists, or, if not, the
variable name. If xy.ticks is FALSE, no ylab is displayed.
Customize these and related parameters with parameters such as
lab.color from the style function.
Label for the title of the graph. If the corresponding variable labels exist, then the title is set by default from the corresponding variable labels.
Sub-title of graph, below xlab.
Two-element vector -- x-axis label, y-axis label -- adjusts the position of the axis labels in approximate inches. + values move the labels away from plot edge. Not applicable to Trellis graphics.
Four-element vector -- top, right, bottom and left -- adjusts the margins of the plotted figure in approximate inches. + values move the corresponding margin away from plot edge. Not applicable to Trellis graphics.
Degrees that the axis values for the value labels on
the x-axis are rotated, usually to accommodate longer values,
typically used in conjunction with offset. When equal 90
the value labels are perpendicular to the x-axis and a different
algorithm places the labels so that offset is not needed.
Degrees that the axis values for the value labels on
the y-axis are rotated, usually to accommodate longer values,
typically used in conjunction with offset.
The amount of spacing between the axis values and the axis. Default is 0.5. Larger values such as 1.0 are used to create space for the label when longer axis value names are rotated.
Flag that indicates if tick marks and associated axis values
on the axes are to be displayed. To rotate the axis values, use
rotate.x, rotate.y, and offset from the
style function.
Labels for the x-axis on the graph to override
existing data values, including factor levels. If the variable is a
factor and value.labels is not specified (is NULL), then the
value.labels are set to the factor levels with each space replaced by
a new line character. If x and y-axes have the same scale,
they also apply to the y-axis.
Maximum size of labels for the values of a categorical variable. Not a literal maximum as preserving unique values may require a larger number of characters than specified.
Origin of x-axis. Particularly useful for plots of
count, etc, where the origin is zero by default, but can
be modified. Otherwise the origin of the plot is based on the minimum
value of x.
Draw one or more objects, text or a geometric figures,
on the plot.
Possible values are any text to be written, the first argument, which is
"text", or, to indicate a figure,
"rect" (rectangle), "line", "arrow",
"v.line" (vertical line), and "h.line" (horizontal line).
The value "means" is short-hand for vertical and horizontal lines
at the respective means. Does not apply to Trellis graphics.
Customize with parameters such as add.fill and add.color
from the style function.
First x coordinate to be considered for each object, can be
"mean.x". Not used for "h.line".
First y coordinate to be considered for each object, can be
"mean.y". Not used for"v.line".
Second x coordinate to be considered for each object, can be
"mean.x". Only used for "rect", "line" and
arrow.
Second y coordinate to be considered for each object, can be
"mean.y". Only used for "rect", "line" and
arrow.
For a two-variable scatterplot, if TRUE,
automatically add the 0.95 data ellipse,
labeling of outliers beyond a Mahalanobis distance of 6 from the
ellipse center, the best-fitting least squares line of all the data,
the best-fitting least squares line of the regular data without the
outliers, and a horizontal and vertical line to represent the mean of
each of the two variables.
Determines if to check for existing data frame and
specified variables. By default is TRUE
unless the shiny package is loaded then set to FALSE so
that Shiny will run. Needs to be set to FALSE if using
the pipe %\>% notation.
Number of significant digits for each of the displayed summary statistics.
If set to TRUE, no text output. Can change system default
with style function.
If TRUE, the default, then generate the plot.
Width of the plot window in inches, defaults to 5 except in RStudio to maintain an approximate square plotting area.
Height of the plot window in inches, defaults to 4.5 except for 1-D scatterplots and when in RStudio.
Indicate to direct pdf graphics to the specified name of the pdf file.
Function call. Used with knitr to pass the function call when
obtained from the abbreviated function call sp.
Other parameter values for non-Trellis graphics as defined by and
processed by standard R functions plot and par,
including
xlim and ylim for setting the range of the x and
y-axes
cex.main for the size of the title
col.main for the color of the title
cex for the size of the axis value labels
sub and col.sub for a subtitle and its color
The output can optionally be saved into an R object, otherwise it simply appears in the console. The output here is just for the outlier analysis of the two-variable scatterplot with continuous variables. The outlier identification must be activated for the analysis, such as from parameter MD.cut.
READABLE OUTPUT codeout_outlier: Mahalanobis Distance of each outlier.
STATISTICS codeoutliers_indices: Location of the outliers in the x and y vectors.
VARIABLES and TRELLIS PLOTS
There is at least one primary variable, x, which defines the coordinate system for plotting in terms of the x-axis, the horizontal axis. Plots may also specify a second primary variable, y, which defines the y-axis of the coordinate system. One of these primary variables may be a vector. The simplest plot is from the specification of only one or two primary variables, each as a single variable, which generates a single scatterplot of either one or two variables, necessarily on a single plot, called a panel, defined by a single x-axis and usually a single y-axis.
For numeric primary variables, a single panel may also contain multiple plots of two types. Form the first type from subsets of observations (rows of data) based on values of a categorical variable. Specify this plot with the by parameter, which identifies the grouping variable to generate a scatterplot of the primary variables for each of its levels. The points for each group are plotted with a different shape and/or color. By default, the colors vary, though to maintain the color scheme, if there are only two levels of the grouping variable, the points for one level are filled with the current theme color and the points for the second level are plotted with transparent interiors.
Or, obtain multiple scatterplots on the same panel with multiple numeric x-variables, or multiple y-variables. To obtain this graph, specify one of the primary variables as a vector of multiple variables.
Trellis graphics, from Deepayan Sarkar's (2009) lattice package, may be implemented in which multiple panels for one numeric x-variable and one numeric y-variable are displayed according to the levels of one or two categorical variables, called conditioning variables. A variable specified with by is a conditioning variable that results in a Trellis plot, the scatterplot of x and y produced at each level of the by1 variable. The inclusion of a second conditioning variable, by2, results in a separate scatterplot panel for each combination of cross-classified values of both by1 and by2. A grouping variable according to by may also be specified, which is then applied to each panel.
Control the panel dimensions and the overall size of the Trellis plot with the following parameters: width and height for the physical dimensions of the plot window, n.row and n.col for the number of rows and columns of panels, and aspect for the ratio of the height to the width of each panel. The plot window is the standard graphics window that displays on the screen, or it can be specified as a pdf file with the pdf.file parameter.
CATEGORICAL VARIABLES
Conceptually, there are continuous variables and categorical variables. Categorical variables have relatively few unique data values. However, categorical variables can be defined with non-numeric values, but also with numeric values, such as responses to a five-point Likert scale from Strongly Disagree to Strongly Agree, with responses coded 1 to 5. The three by --variables -- by1, by2 and by -- only apply to graphs created with numeric x and/or y variables, continuous or categorical.
The standard and most general way to define a categorical variable is as an R factor, such as created with the lessR doFactors function. lessR provides the option to define an integer variable with equally spaced values as categorical based on the value of n.cat, which can be set locally or globally with the style function. For example, for a variable with data values from 5-point Likert scale, a value of n.cat of 5 will define the define the variable as categorical. The default value is 0. To explicitly analyze the values as categorical, set n.cat to a value larger than 0, at least the size of the number of unique integer values. Can also annotate a graph of the values of an integer categorical variable with value.labels option.
A scatterplot of Likert type data is problematic because there are so few possibilities for points in the scatterplot. For example, for a scatterplot of two five-point Likert response data, there are only 26 possible paired values to plot, so most of the plotted points overlap with others. In this situation, that is, when a single variable or two variables with Likert response scales are specified, a bubble plot is automatically provided, with the size of each point relative to the joint frequency of the paired data values. To request a sunflower plot in lieu of the bubble plot, set the shape to "sunflower".
DATA
The default input data frame is mydata. Specify another name with the data option. Regardless of its name, the data frame need not be attached to reference the variables directly by its name, that is, no need to invoke the mydata$name notation. The referenced variables can be in the data frame and/or the user's workspace, the global environment.
The data values themselves can be plotted, or for a single variable, counts or proportions can be plotted on the y-axis. For a categorical x-variable paired with a continuous variable, means and other statistics can be plotted at each level of the x-variable. If x is continuous, it is binned first, with the standard Histogram binning parameters available, such as bin.width, to override default values. The values parameter sets the values to plot, with data the default. By default, the connecting line segments are provided, so a frequency polygon results. Turn off the lines by setting lwd=0.
The rows parameter subsets rows (cases) of the input data frame according to a logical expression. Use the standard R operators for logical statements as described in Logic such as & for and, | for or and ! for not, and use the standard R relational operators as described in Comparison such as == for logical equality != for not equals, and > for greater than. See the Examples.
VALUE LABELS
The value labels for each axis can be over-ridden from their values in the data to user supplied values with the value.labels option. This option is particularly useful for Likert-style data coded as integers. Then, for example, a 0 in the data can be mapped into a "Strongly Disagree" on the plot. These value labels apply to integer categorical variables, and also to factor variables. To enhance the readability of the labels on the graph, any blanks in a value label translate into a new line in the resulting plot. Blanks are also transformed as such for the labels of factor variables.
However, the lessR function doFactors allows for the easy creation of factors, one variable or a vector of variables, in a single statement, and is generally recommended as the method for providing value labels for the variables.
VARIABLE LABELS
Although standard R does not provide for variable labels, lessR can store the labels in the data frame with the data, obtained from the Read function or VariableLabels. If variable labels exist, then the corresponding variable label is by default listed as the label for the corresponding axis and on the text output.
ONE VARIABLE PLOT
The one variable plot of one continuous variable generates either a violin/box/scatterplot (VBS plot), or a run chart with run=TRUE, or x can be an R time series variable for a time series chart. For the box plot,
for gray scale output potential outliers are plotted with squares and outliers are plotted with diamonds, otherwise shades of red are used to highlight outliers. The default definition of outliers is based on the standard boxplot rule of values more than 1.5 IQR's from the box. The definition of outliers may be adjusted (Hubert and Vandervieren, 2008), such that the whiskers are computed from the medcouple index of skewness (Brys, Hubert, & Struyf, 2004).
The plot can also be obtained as a bubble plot of frequencies for a categorical variable.
TWO VARIABLE PLOT
When two variables are specified to plot, by default if the values of the first variable, x, are unsorted, or if there are unequal intervals between adjacent values, or if there is missing data for either variable, a scatterplot is produced from a call to the standard R plot function. By default, sorted values with equal intervals between adjacent values of the first of the two specified variables yields a function plot if there is no missing data for either variable, that is, a call to the standard R plot function with type="l", which connects each adjacent pair of points with a line segment.
Specifying multiple, continuous x-variables against a single y variable, or vice versa, results in multiple plots on the same graph. The color of the points of the second variable is the same as that of the first variable, but with a transparent fill. For more than two x-variables, multiple colors are displayed, one for each x-variable.
BUBBLE PLOT FREQUENCY MATRIX (BPFM)
Multiple categorical variables for x may be specified in the absence of a y variable. A bubble plot results that illustrates the frequency of each response for each of the variables in a common figure in which the x-axis contains all of the unique labels for all of the variables plotted. Each line of information, the bubbles and counts for a single variable, replaces the standard bar chart in a more compact display. Usually the most meaningful when each variable in the matrix has the same response categories, that is, levels, such as for a set of shared Likert scales. The BPFM is considerably condensed presentation of frequencies for a set of variables than are the corresponding bar charts.
SCATTERPLOT MATRIX
A single vector of continuous variables specified as x, with no y-variable, generates a scatterplot matrix of the specified variable. A continuous variable is defined as a numeric variable with more than n.cat unique responses. To force an item with a small number of unique responses, such as from a 5-pt Likert scale, to be treated as continuous, set n.cat to a number lower than 5, such as n.cat=0 in the function call.
The scatterplot matrix is displayed according to the current color theme. Specific colors such as fill, color, etc. can also be provided. The upper triangle shows the correlation coefficient, and the lower triangle each corresponding scatterplot, with, by default, the non-linear loess best fit line. The code fit option can be used to provide the linear least squares line instead, along with the corresponding fit.color for the color of the fit line.
SIZE VARIABLE
A variable specified with size= is a numerical variable that activates a bubble plot in which the size of each bubble is determined by the value of the corresponding value of size, which can be a variable or a constant.
To explicitly vary the shapes, use shape and a list of shape values in the standard R form with the c function to combine a list of values, one specified shape for each group, as shown in the examples. To explicitly vary the colors, use fill, such as with R standard color names. If fill is specified without shape, then colors are varied, but not shapes. To vary both shapes and colors, specify values for both options, always with one shape or color specified for each level of the by variable.
Shapes beyond the standard list of named shapes, such as "circle", are also available as single characters. Any single letter, uppercase or lowercase, any single digit, and the characters "+", "*" and "#" are available, as illustrated in the examples. In the use of shape, either use standard named shapes, or individual characters, but not both in a single specification.
SCATTERPLOT ELLIPSE
For a scatterplot of two numeric variables, the ellipse=TRUE option draws the .95 data ellipse as computed by the ellipse function, written by Duncan Murdoch and E. D. Chow, from the ellipse package. The axes are automatically lengthened to provide space for the entire ellipse that extends beyond the maximum and minimum data values. The specific level of the ellipse can be specified with a numerical value in the form of a proportion. Multiple numerical values of ellipse may also be specified to obtain multiple ellipses.
BOXPLOTS
For a single variable the preferred plot is the integrated violin/box/scatter plot or VBS plot. Only the violin or box plot can be obtained with the corresponding aliases ViolinPlot and BoxPlot, or by setting vbs.plot to "v" or "b". To view a box plot of a continuous variable (Y) across the levels of a categorical variable (X), either as part of the full VBS plot, or by itself, there are two possibilities:
1. Plot(Y,X) or BoxPlot(Y, X)
2. Plot(Y, by1=X) or BoxPlot(Y, by1=X)
Both styles produce the same information. What differs is the color scheme.
The first possibility places the multiple box plots on a single pane and also, for the default color scheme "colors", displays the sequence of box plots with the default qualitative color palette from the lessR function getColors.
All colors are displayed at the same level of gray-scale saturation and brightness to avoid perceptual bias. BarChart and PieChart use the same default colors as well.
The second possibility with by1 produces the different box plots on a separate panel, that is, a Trellis chart. These box plots are displayed with a single hue, the first color, blue, in the default qualitative sequence.
TIME CHARTS
Specifying one or more x-variables with no y-variables, and run=TRUE plots the x-variables in a run chart. The values of the specified x-variable are plotted on the y-axis, with Index on the x-axis. Index is the ordinal position of each data value, from 1 to the number of values.
If the specified x-variable is of type Date, or is a time series, a time series plot is generated for each specified variable. If a formal R time-series, univariate or multivariate, specify as the x-variable. Or, specify the x-variable of type Date, and then specify the y-variable as one or more time series to plot. The y-variable can be formatted as tidy data with all the values in a single column, or as wide-formatted data with the time-series variables in separate columns.
2-D KERNEL DENSITY
With smooth=TRUE, the R function smoothScatter is invoked according to the current color theme. Useful for very large data sets. The smooth.points parameter plots points from the s of the lowest density. The smooth.bins parameter specifies the number of bins in both directions for the density estimation. The smooth.trans parameter specifies the exponent in the function that maps the density scale to the color scale to allow customization of the intensity of the plotted gradient colors. Higher values result in less color saturation, de-emphasizing points from regions of lessor density. These parameters are respectively passed directly to the smoothScatter nrpoints, nbin and transformation parameters. Grid lines are turned off, but can be displayed by setting the grid.color parameter.
COLORS
A color theme for all the colors can be chosen for a specific plot with the colors option with the lessR function style. The default color theme is "lightbronze". A gray scale is available with "gray", and other themes are available as explained in style, such as "sienna" and "darkred". Use the option style(sub.theme="black") for a black background and partial transparency of plotted colors.
Colors can also be changed for individual aspects of a scatterplot as well with the style function. To provide a warmer tone by slightly enhancing red, try a background color such as panel.fill="snow". Obtain a very light gray with panel.fill="gray99". To darken the background gray, try panel.fill="gray97" or lower numbers. See the lessR function showColors, which provides an example of all available named R colors with their RGB values.
For the color options, such as violin.color, the value of "off" is the same as "transparent".
ANNOTATIONS
Use the add and related parameters to annotate the plot with text and/or geometric figures. Each object is placed according from one to four corresponding coordinates, the required coordinates to plot that object, as shown in the following table. x-coordinates may have the value of "mean.x" and y-coordinates may have the value of "mean.y".
| Value | Object | Required Coordinates |
| ----------- | ------------------- | ---------------- |
| text | text | x1, x2 |
"rect" |
rectangle | x1, y1, x2, y2 |
"line" |
line segment | x1, y1, x2, y2 |
"arrow" |
arrow | x1, y1, x2, y2 |
"v.line" |
vertical line | x1 |
"h.line" |
horizontal line | y1 |
"means" |
horiz, vert lines | |
| ----------- | ------------------- | ---------------- |
The value of add specifies the object. For a single object, enter a single value. Then specify the value of the needed corresponding coordinates, as specified in the above table. For multiple placements of that object, specify vectors of corresponding coordinates. To annotate multiple objects, specify multiple values for add as a vector. Then list the corresponding coordinates, for up to each of four coordinates, in the order of the objects listed in add. See the examples for illustrations.
Can also specify vectors of different properties, such as add.color. That is, different objects can be different colors, different transparency levels, etc.
PDF OUTPUT
To obtain pdf output, use the pdf.file option, perhaps with the optional width and height options. These files are written to the default working directory, which can be explicitly specified with the R setwd function.
ADDITIONAL OPTIONS
Commonly used graphical parameters that are available to the standard R function plot are also generally available to Plot, such as:
Settings for main- and sub-title and axis annotation, see title and par.
Title of the graph, see title.
The limits of the plot on the x-axis, expressed as c(x1,x2), where x1 and x2 are the limits. Note that x1 > x2 is allowed and leads to a reversed axis.
The limits of the plot on the y-axis.
ONLY VARIABLES ARE REFERENCED
A referenced variable in a lessR function can only be a variable name. This referenced variable must exist in either the referenced data frame, such as the default mydata, or in the user's workspace, more formally called the global environment. That is, expressions cannot be directly evaluated. For example:
> Plot(rnorm(50), rnorm(50)) # does NOT work
Instead, do the following:
> X <- rnorm(50) # create vector X in user workspace
> Y <- rnorm(50) # create vector Y in user workspace
> Plot(X,Y) # directly reference X and YBrys, G., Hubert, M., & Struyf, A. (2004). A robust measure of skewness. Journal of Computational and Graphical Statistics, 13(4), 996-1017.
Murdoch, D, and Chow, E. D. (2013). ellipse function from the ellipse package package.
Gerbing, D. W. (2014). R Data Analysis without Programming, Chapter 8, NY: Routledge.
Hubert, M. and Vandervieren, E. (2008). An adjusted boxplot for skewed distributions, Computational Statistics and Data Analysis 52, 51865201.
Sarkar, Deepayan (2008) Lattice: Multivariate Data Visualization with R, Springer. http://lmdvr.r-forge.r-project.org/
plot, stripchart, title, par, loess, Correlation, style.
# NOT RUN {
# read the data
mydata <- rd("Employee", in.lessR=TRUE, quiet=TRUE)
mydata <- Subset(random=.6, quiet=TRUE) # less computationally intensive
#---------------------------------------------------
# traditional scatterplot with two numeric variables
#---------------------------------------------------
# scatterplot with all defaults
Plot(Years, Salary)
# or use abbreviation sp in place of Plot
# or use full expression ScatterPlot in place of Plot
# maximum information, minimum input: scatterplot +
# means, outliers, ellipse, least-squares lines with and w/o outliers
Plot(Years, Salary, auto=TRUE)
# }
# NOT RUN {
# just males employed more than 5 years
Plot(Years, Salary, rows=(Gender=="M" & Years > 5))
# plot 0.95 data ellipse with the points identified that represent
# outliers defined by a Mahalanobis Distance larger than 6
# save outliers into R object out, then remove from mydata
mydata[1, "Salary"] <- 200000
out <- Plot(Years, Salary, ellipse=0.95, MD.cut=6)
mydata <- mydata[-out$outlier_indices,]
# new shape and point size, no grid or background color
# then put style back to default
style(panel.fill="powderblue", grid.color="off")
Plot(Years, Salary, size=2, shape="diamond")
style()
# translucent data ellipses without points or edges
# show the idealized joint distribution for bivariate normality
style(ellipse.color="off")
Plot(Years, Salary, size=0, ellipse=seq(.1,.9,.10))
style()
# bubble plot with size determined by the value of Pre
# display the value for the bubbles with values of min, median and max
Plot(Years, Salary, size=Pre, size.cut=3)
# variables in a data frame not the default mydata
# plot 0.6 and 0.9 data ellipses with partially transparent points
# change color theme to gold with black background
style("gold", sub.theme="black")
Plot(eruptions, waiting, trans=.5, ellipse=seq(.6,.9), data=faithful)
# scatterplot with two x-variables, plotted against Salary
# define a new style, then back to default
style(window.fill=rgb(247,242,230, maxColorValue=255),
panel.fill="off", panel.color="off", pt.fill="black", trans=0,
lab.color="black", axis.text.color="black",
axis.y.color="off", grid.x.color="off", grid.y.color="black",
grid.lty="dotted", grid.lwd=1)
Plot(c(Pre, Post), Salary)
style()
# increase span (smoothing) from default of .7 to 1.25
# span is a loess parameter, which generates a caution that can be
# ignored that it is not a graphical parameter -- we know that
# display confidence intervals about best-fit line at
# 0.95 confidence level
Plot(Years, Salary, fit="loess", span=1.25)
# 2-D kernel density (more useful for larger sample sizes)
Plot(Years, Salary, smooth=TRUE)
# }
# NOT RUN {
#------------------------------------------------------
# scatterplot matrix from a vector of numeric variables
#------------------------------------------------------
# with least squares fit line
Plot(c(Salary, Years, Pre), fit="lm")
#--------------------------------------------------------------
# Trellis graphics and by for groups with two numeric variables
#--------------------------------------------------------------
# Trellis plot with condition on 1-variable
Plot(Years, Salary, by1=Dept)
# }
# NOT RUN {
# all three by variables
Plot(Years, Salary, by1=Dept, by2=Gender, by=Plan)
# vary both shape and color with a least-squares fit line for each group
style(color=c("darkgreen", "brown"))
Plot(Years, Salary, by1=Gender, fit="lm", shape=c("F","M"), size=.8)
style("gray")
# compare the men and women Salary according to Years worked
# with an ellipse for each group
Plot(Years, Salary, by=Gender, ellipse=.50)
# }
# NOT RUN {
#--------------------------------------------------
# analysis of a single numeric variable (or vector)
#--------------------------------------------------
# One continuous variable
# -----------------------
# integrated Violin/Box/Scatterplot, a VBS plot
Plot(Salary)
# }
# NOT RUN {
# by variable, different colors for different values of the variable
# all on one panel
Plot(Salary, by=Dept)
# large sample size
x <- rnorm(10000)
Plot(x)
# custom colors for outliers, which might not appear in this subset data
style(out.fill="hotpink", out2.fill="purple")
Plot(Salary)
style()
# no violin plot or scatterplot, just a boxplot
Plot(Salary, vbs.plot="b")
# or, the same with the mnemonic
BoxPlot(Salary)
# two related displays of box plots for different levels of a
# categorical variable
# 1. With default theme of "colors" display box plots with different colors
BoxPlot(Salary, Dept)
# 2. Display box plots with the same hue (blue by default)
BoxPlot(Salary, by1=Dept)
# binned values to plot counts
# ----------------------------
# bin the values of Salary to plot counts as a frequency polygon
# the counts are plotted as points instead of the data
Plot(Salary, values="count") # bin the values
# time charts
#------------
# run chart, with fill area
Plot(Salary, run=TRUE, area=TRUE)
# two run charts in same plot
# or could do a multivariate time series
Plot(c(Pre, Post), run=TRUE)
# Trellis graphics run chart with custom line width, no points
Plot(Salary, run=TRUE, by1=Gender, lwd=3, size=0)
# daily time series plot
# create the daily time series from R built-in data set airquality
oz.ts <- ts(airquality$Ozone, start=c(1973, 121), frequency=365)
Plot(oz.ts)
# multiple time series plotted from dates and stacked
# black background with translucent areas, then reset theme to default
style(sub.theme="black", color="steelblue2", trans=.55,
window.fill="gray10", grid.color="gray25")
date <- seq(as.Date("2013/1/1"), as.Date("2016/1/1"), by="quarter")
x1 <- rnorm(13, 100, 15)
x2 <- rnorm(13, 100, 15)
x3 <- rnorm(13, 100, 15)
df <- data.frame(date, x1, x2, x3)
rm(date); rm(x1); rm(x2); rm(x3)
Plot(date, x1:x3, data=df)
style()
# trigger a time series with a Date variable specified first
# stock prices for three companies by month: Apple, IBM, Intel
mydata <- rd("StockPrice", in.lessR=TRUE)
# only plot Apple
Plot(date, Price, rows=(Company=="Apple"))
# Trellis plots, one for each company
Plot(date, Price, by1=Company, n.col=1)
# all three plots on the same panel, three shades of blue
Plot(date, Price, by=Company, color="blues")
# }
# NOT RUN {
#------------------------------------------
# analysis of a single categorical variable
#------------------------------------------
mydata <- rd("Employee", in.lessR=TRUE)
# default 1-D bubble plot
# frequency plot, replaces bar chart
Plot(Dept)
# }
# NOT RUN {
# abbreviated category labels
Plot(Dept, label.max=2)
# plot of frequencies for each category (level), replaces bar chart
Plot(Dept, values="count")
#----------------------------------------------------
# scatterplot of numeric against categorical variable
#----------------------------------------------------
# generate a chart with the plotted mean of each level
# rotate x-axis labels and then offset from the axis
style(rotate.x=45, offset=1)
Plot(Dept, Salary)
style()
#-------------------
# Cleveland dot plot
#-------------------
# row.names on the y-axis
Plot(Salary, row.names)
# standard scatterplot
Plot(Salary, row.names, sort.yx=FALSE, segments.y=FALSE)
# Cleveland dot plot with two x-variables
Plot(c(Pre, Post), row.names)
#------------
# annotations
#------------
# add text at the one location specified by x1 and x2
Plot(Years, Salary, add="Hi There", x1=12, y1=80000)
# add text at three different specified locations
Plot(Years, Salary, add="Hi", x1=c(12, 16, 18), y1=c(80000, 100000, 60000))
# add three different text blocks at three different specified locations
Plot(Years, Salary, add=c("Hi", "Bye", "Wow"), x1=c(12, 16, 18),
y1=c(80000, 100000, 60000))
# add an 0.95 data ellipse and horizontal and vertical lines through the
# respective means
Plot(Years, Salary, ellipse=0.95, add=c("v.line", "h.line"),
x1="mean.x", y1="mean.y")
# can be done also with the following short-hand
Plot(Years, Salary, ellipse=0.95, add=c("means"))
# a rectangle requires two points, four coordinates, <x1,y1> and <x2,y2>
style(add.trans=.8, add.fill="gold", add.color="gold4", add.lwd=0.5)
Plot(Years, Salary, add="rect", x1=12, y1=80000, x2=16, y2=115000)
# the first object, a rectangle, requires all four coordinates
# the vertical line at x=2 requires only an x1 coordinate, listed 2nd
Plot(Years, Salary, add=c("rect", "v.line"), x1=c(10, 2),
y1=80000, x2=12, y2=115000)
# two different rectangles with different locations, fill colors and translucence
style(add.fill=c("gold3", "green"), add.trans=c(.8,.4))
Plot(Years, Salary, add=c("rect", "rect"),
x1=c(10, 2), y1=c(60000, 45000), x2=c(12, 75000), y2=c(80000, 55000))
# }
# NOT RUN {
#----------------------------------------------------
# analysis of two categorical variables (Likert data)
#----------------------------------------------------
mydata <- rd("Mach4", in.lessR=TRUE, quiet=TRUE) # Likert data, 0 to 5
# size of each plotted point (bubble) depends on its joint frequency
# triggered by default when replication of joint values and
# less than 9 unique data values for each
# n.cat=6 means treat responses as categorical for up to 6 equally-spaced
# integer values
Plot(m06, m07, n.cat=6)
# use value labels for the integer values, modify color options
LikertCats <- c("Strongly Disagree", "Disagree", "Slightly Disagree",
"Slightly Agree", "Agree", "Strongly Agree")
style(fill="powderblue", color="blue", bubble.text="darkred")
Plot(m06, m07, value.labels=LikertCats, n.cat=6)
style() # reset theme
# }
# NOT RUN {
# proportions within each level of the other variable
Plot(m06, m07, proportion=TRUE, n.cat=6)
# get correlation analysis instead of cross-tab analysis
# rely upon the default value of n.cat=0 so that integer
# valued variables are analyzed as numerical
Plot(m06, m07)
#-----------------------------
# Bubble Plot Frequency Matrix
#-----------------------------
# applies to categorical variables, since Mach IV Likert items
# are 0 to 5 integer values, set n.cat to indicate the
# numeric values represent categories
Plot(c(m06,m07,m09,m10), value.labels=LikertCats, n.cat=6)
#---------------
# function curve
#---------------
x <- seq(10,50,by=2)
y1 <- sqrt(x)
y2 <- x**.33
# x is sorted with equal intervals so run chart by default
Plot(x, y1)
# multiple plots from variable vectors need to have the variables
# in a data frame
mydata <- data.frame(x, y1, y2)
# if variables are in the user workspace and in a data frame
# with the same names, the user workspace versions are used,
# which do not work with vectors of variables, so remove
rm(x); rm(y1); rm(y2)
Plot(x, c(y1, y2))
#-----------
# modern art
#-----------
clr <- colors() # get list of color names
color0 <- clr[sample(1:length(clr), size=1)]
clr <- clr[-(153:353)] # get rid of most of the grays
n <- sample(5:30, size=1)
x <- rnorm(n)
y <- rnorm(n)
color1 <- clr[sample(1:length(clr), size=1)]
color2 <- clr[sample(1:length(clr), size=1)]
style(window.fill=color0, area.fill=color1, color=color2)
Plot(x, y, run=TRUE, area=TRUE,
xy.ticks=FALSE, main="Modern Art", xlab="", ylab="",
cex.main=2, col.main="lightsteelblue", n.cat=0, center.line="off")
style() # reset style to default
# }
Run the code above in your browser using DataLab