R Graphics

1
R Graphics

• Dr. Yan Liu
• Department of Biomedical, Industrial and Human
  Factors Engineering
• Wright State University

2
Introduction to R

• What is R
• A free open-source system for statistical
  computation and graphics
• Consists of a language (called R) plus a run-time
  environment with graphics, debugger, access to
  certain system functions, and the ability to run
  programs stored in script files
• Influenced by S language, developed by Becker,
  Chamber, and Wilks at Bell Laboratories
• S is a very high level language and an
  environment for data analysis and graphics
• S-Plus, a commercial tool
• Initially written by Ross Ihaka and Robert
  Gentleman at the Department of Statistics of the
  University of Auckland in Auckland, New Zealand
• Possible for the user to interface to procedures
  written in the C, C, or FORTRAN languages for
  efficiency
• Main Website of R
• Http//cran.r-project.org/ (download Linux,
  MacOS X, and Windows)

3
Start Up

• Two Alternatives to Run Commands in R
• Command window (R Console window)
• Script file (File gtgt New script)
• Highlight the commands in the script file window
  and click the run line or selection button

run line or selection button
Script File Window
Command Window
4
Read Files

• Read in Data from an External File
• Parameters file the name and directory of
  the file from which the data are to be read
• header T the first
  row in the table of data includes the attribute
  names of the data
• sep the field
  separator character. (\t means separation by
  tab other common separators
  include , and )
• na.strings
  specify the missing characters, which is NA by
  default
• read.csv( ) is identical to read.table except for
  the defaults. It is intended for reading comma
  separated value files (.csv)

5
Characteristics of Dataset
gt names(auto) returns the attribute names of
auto dataset gt str(auto) returns the attribute
names of auto dataset and a short description
of each attribute and the dataset
6
Basic Attribute Types

• Numeric
• Real numbers
• Integer
• Logical
• Binary true or false
• Character/Strings
• e.g. red, green
• Factor
• Categorical attribute whose values are stored as
  a vector of integers in the range 1... k
  (where k is the number of unique values in the
  nominal variable)
• e.g. In attribute country 1 - USA, 2 European,
  3 Japan
• An ordered factor is used to represent an ordinal
  variable
• e.g. In attribute size 1 - small, 2 medium, 3
  large

7
Convert Attribute Type

• as. numeric(x)
• Convert an attribute to numeric
• as. integer(x)
• Convert an attribute to integer
• as. factor(x)
• Convert an attribute to factor
• toString(x, width)
• Convert an attribute to characters/strings

8
R Objects

• Scalar
• A single value
• Vector
• A one-dimensional array of arbitrary length
• gt c(2, 3, 5, 2, 7, 1)
• gt 310
• gt c(Canberra, Sydney, Newcastle)
• All elements of the vector must be of the same
  type (e.g. numerical, character, etc.)
• Subsets of the vector may be referenced
• gt x lt- c(2, 3, 5, 2, 7, 1)
• gt xc(2,4) extract elements 2 and 4 of x
• gt x-c(2,4) extract elements of x except
  elements 2 and 4

9
R Objects (Cont.)

• Matrix
• A two-dimensional array with an arbitrary number
  of rows and columns
• All elements of the matrix must be of the same
  type
• Subsets of the matrix may be referenced
• Individual rows and columns of the matrix may be
  handled as vectors

The first two elements at the 1st row
The elements at the first two columns
10
R Objects (Cont.)

• Array
• As a matrix, but of arbitrary dimension
• Data Frame
• A dataset with rows (representing data records)
  and columns (representing attributes)
• May be handled similarly to a matrix
• Individual columns of the data frame may be
  handled as vectors

11
(No Transcript)
12
R Objects (Cont.)

• Function
• R has a vast number of built-in' functions
• e.g. mean( ), plot( ), var( ), etc.
• Users can write their own functions
• List
• An arbitrary collection of other R objects (which
  may include other lists)

• Quit Function q()
• On quitting, R offers the options of saving the
  workspace image, in the file .RData in the
  working directory
• Remove Object Function rm()
• Remove objects that are no longer needed

13
A Simple Scatterplot

• plot (autompg, autohorsepower) produces a
  scatterplot of mpg vs. horsepower of the auto
  dataset /

text(40, 200, Plot of mpg vs. horsepower) adds
the label at the location (40, 200) within the
plot
14
Overview of R Graphics

• Graphics Functions
• High-level functions that produce complete plots
• Some flexibility in the way that the data to be
  plot can be specified
• e.g. plot( )
• Low-level functions that add some outputs to
  existing plots
• e.g. text( )
• Functions for working interactively with
  graphical outputs
• Painters Model
• Graphics output occurs in steps, later output
  obscuring any previous output that it overlaps

15
Traditional Standard Plots
16
Trellis Plots

• Provided through package Lattice
• Embody a number of design principles proposed by
  Bill Cleveland (1987, 2004) that aim to ensure
  effective visualization
• Trellis Display

17

• When there are many overlapping points, we can
  make points semi-transparent to mitigate the
  overlapping issue

Where the color is "RRGGBBAA" and the AA portion
is the opacity/trasparency
18
Special-Purpose Plots

• R provides a set of functions for producing
  graphical output primitives (e.g. lines, text,
  rectangles, polygons, etc.) which users can use
  to create plots with special purposes

19
Graphical Output Formats

• When using R interactively, the result is a plot
  drawn on screen
• Can be saved as a PDF, postcript, or image file
• File gt Save as gt Postcript/PDF/Png (a desired
  format)
• Can produce a file that contains the plot
• Output is directed to a particular output device
  which indicates the output format
• postscript( ) for Adobe PostScript file, pdf( )
  for Adobe PDF file, pictex( ) for LaTex PicTex
  file,
• png ( ) for PNG bitmap file, jpeg( ) for JPEG
  bitmap file, bmp( ) for Window BMP file
• Close a device
• dev.off ( )

A PDF file of the plot will be saved in the
same directory as that of the R workspace
20
Structure of the R Graphics System

• Core Graphics Systems
• Graphics (traditional graphics)
• Grid
• Lattice package is built on Grid

• Graphics Engine Devices
• grDevices package consists of functions that
  provide support for handling colors and font

Structure of the R Graphics System
(Showing the main packages that provide graphics
functions in R. Arrows indicate where one package
builds on the functions in another package)
21
Traditional versus Grid Graphics Systems

• High-Level Functions
• The traditional system, or the graphics package
  built on the top of it, provide the majority of
  the high-level functions currently available in R
• Lattice package, built on the Grid system,
  provides high-level functions
• Low-Level Functions
• Both provide many low-level functions
• Functions for Interaction
• Traditional system provides very limited
  interaction
• Grid system provides functions for interacting
  with graphical outputs
• Editing, extracting, deleting parts of an image
• Graphics Design
• Trellis plots have a better design in terms of
  visually encoding information (based on research
  on human visual perception)

22
Lattice Graphics Model

• Loading Lattice into R

• Lattice Plot Types
• A number of standard plot types (like those in
  the traditional graphics)
• More modern and specialized plots
• A table of comparison of plot functions of
  lattice and traditional graphics systems can be
  downloaded from the course website

• A Lattice graphics function produces an object of
  class trellis which contains description of the
  plot

• Possible to work with the trellis object and
  modify it using the update() function for
  trellis objects

23
(No Transcript)
24
Trellis Display xyplot

• xyplot(yxg1,g2,, data, ) produces a
  scatterplot of y (on vertical axis) versus x (on
  horizontal axis) conditioning on g1, g2,
• Create shingles for conditioning variables with
  continuous values
• A shingle is a data structure that consists of a
  numeric vector along with some possibly
  overlapping intervals
• equal.count(x, number, overlap)
• Create a shingle that consists of intervals with
  (almost) the same number of data records
• x the variable to be shingled number the
  number of intervals overlap the overlapping
  between successive intervals (as proportion to
  the number of records in each interval)

25
(No Transcript)
26
Trellis Display 3D Scatterplot

• cloud(zxyg1,g2,, data, ) produces a 3D
  scatterplot of z (on vertical axis) versus x and
  y (on horizontal grid) conditioning on g1, g2,

27
Parallel Coordinates

• Parallel(x, data, ) produces a parallel
  coordinates of data frame x

28
Rotate Plot
29

• Parallel(xg1,g2,, data, ) produces a parallel
  coordinates of data frame x conditioning on g1,
  g2,

30
R Formula

• The first argument to the lattice plotting
  functions is usually an R formula
• Common Types
• yx plots variable y (on the vertical axis)
  against variable x (on the horizontal axis)
• x used in plots of one variable x or parallel
  coordinates of a data frame (matrix) x
• zyx plots variable z against x and y (which
  are on the base grid)
• y1y2x plots both variable y1 and variable y2
  against x

31
(No Transcript)
32
Arranging Lattice Plots

• Arrangement of Panels and Strips in a Single
  Lattice Plot
• layout(mat, )
• mat a matrix object with up to 3 dimensions,
  specifying the number of the columns, rows, and
  pages
• aspect argument specifies the aspect ratio
  (height divided by width) for the panels
• aspectfill by default which means to make the
  panel to fill the available space
• aspect xy means the aspect ratio is
  calculated to satisfy the banking to 45

 "The aspect ratio is vital because it has a
large impact on our ability to judge rate of
change. A number of studies in visual perception
have shown that our ability to judge the relative
slopes of line segments on a graph is maximized
when the absolute values of the orientations of
the segments are centered on 45 degrees. Bill
Cleveland (http//stat.bell-labs.com/project/trell
is/interview.html )
33
(No Transcript)
34

• Arrangement of Several Lattice Plots on a Single
  Page
• First, create a trellis object for each lattice
  plot
• Then, call print( ), supplying arguments to
  specify the position of each plot

35
Traditional Plots of One or Two Variables

• plot( ) produces scatterplots

36
Traditional Plots of One or Two Variables (Cont.)

• Specify data to be plot in plot( )

37
Traditional 3D Plots

• persp(x, y, z, ) produces 3D surfaces with x and
  y as the base coordinates and z is a function of
  x and y

38
Traditional 3D Plots (Cont.)

• symbols(x, y, circles, squares, rectangles,
  stars, thermometers, boxplots, ) uses one of the
  six symbols to represent the third variable

39
Traditional Multivariate Plots

• pairs(x, ) produces a scatterplot matrix of x
  (a matrix or data frame)

40
Traditional Multivariate Plots (Cont.)

• stars(x, ) produces a star plot of x

41
Getting Help

• Every R function and dataset has online help
  associated with it, using help( )
• help(help) gives instructions on how to use help(
  )