Introduction to R

1
Introduction to R

• Summer session Lecture 3
• Brian Healy

2
Outline

• Discussion of R
• Importing and changing data
• Creating your own data
• Summary statistics / graphs
• Tests for normality

3
What is R?

• Statistical computer language similar to S-plus
• Has many built-in statistical functions
• Easy to build your own functions (similar to SAS
  macros)
• Good graphic displays
• Extensive help files

4
Strengths

• Many built-in functions
• Can get other functions from the internet by
  downloading libraries
• Relatively easy data manipulations

Weaknesses

• Not as commonly used by non-statisticians
• Many datasets already in SAS form

5
Starting R

• Windows / HSPH computers
• Open using the start menu or g drive
• Unix / Telnet in HSPH from home
• Open using R1.9 command
• Unix specific commands will be discussed at the
  end of the session

6
Writing R code

• Can input lines one at a time into R
• Can write many lines of code in a text editor and
  run all at once
• Using Windows version, simply paste the commands
  into R
• Using Unix version, save the commands and run in
  batch mode

7
Types of commands

• Defining variables
• Inputting data
• Using built-in functions
• Using the help menu and notation
• ?functionname, help.search(functionname)
• Writing your own functions

8
Language layout

• Three types of statement
• expression it is evaluated, printed, and the
  value is lost (35)
• assignment passes the value to a variable but
  the result is not printed automatically
  (outlt-35)
• comment (This is a comment)

9
Naming conventions

• Any roman letters, digits, and . (non-initial
  position)
• Avoid using system names c, q, s, t, C, D, F, I,
  T, diff, mean, pi, range, rank, tree, var
• Hold for variables, data and functions

10
Arithmetic operations and functions

• Most operations in R are similar to Excel and
  calculators
• Basic (add), -(subtract), (multiply),
  /(divide)
• Exponentiation
• Remainder or modulo operator
• Matrix multiplication
• sin(x), cos(x), cosh(x), tan(x), tanh(x),
  acos(x), acosh(x), asin(x), asinh(x), atan(x),
  atan(x,y) atanh(x)
• abs(x), ceiling(x), floor(x)
• exp(x), log(x, baseexp(1)), log10(x), sqrt(x),
  trunc(x) (the next integer closer to zero)
• max(), min()

11
Defining new variables

• Assignment symbol, use lt- (or _)
• Scalars
• scallt-6
• valuelt-7
• Vectors
• veclt-c(0,1,2)
• vec2lt-c(110)
• vec3lt-c(8,6,4,2,10,12,14)
• famnameslt-c("Kate", "Andrew", "Brian")
• Variable names are case sensitive

12
Examples

• Try the following
• 342
• 5scal
• (valuescal)2
• 3vec
• sqrt(vec2)
• trunc(log(vec3))
• What happened in the last three cases?
• How can we assign the result of these to a new
  variable?
• What is the minimum of vec3?

13
Indexing vectors

• To choose individual observations from a vector,
  use vec1
• Try the following
• Find the 3rd value from vec
• List the 4th, 5th, and 6th values from vec2
• Make a new variable that is only the first 2
  values from vec2

14
Matrix

• There are several ways to make a matrix
• To make a 2x3 (2 rows, 3 columns) matrix of 0s
• matlt-matrix(0,2,3)
• To make the following matrix
• mat2lt-rbind(c(71,172),c(73,169),c(69,160),c(65,130
  ))
• mat3lt-cbind(c(71,73,69,65),c(172,169,160,130))
• To make the following matrix
• mat4lt-matrix(vec2,2,5, byrowT)

15
Matrix Examples

• Create the following matrices
• ex1
• ex2 ex3

16
Rows and columns of matrices

• You can pick an individual data point, row or
  column from a matrix
• mat1,2 will give the observation in the first
  row, second column
• What happens when you type
• mat2,2
• mat22,
• mat413,1

17
Lists

• The final form of data is lists
• Type the following code
• ourlistlt-list(v1vec,v2vec2, famfamnames)
• Now, lets look at ourlist
• If there are no names, we can give the elements
  of our list names using
• names(outlist)lt-c(v1, v2, fam)
• To get individual parts of a list we must use the
  sign
• What happens when you type ourlistv1?
• What happens when you type ourlistv11?

18
Data frames

• Very similar to matrices in form
• Different columns can be of different types,
  unlike matrices
• famageslt-c(24, 29, 27) famheightlt- c(64, 73, 71)
• We can put my family names, ages and heights into
  a data frame
• famlt-data.frame(famnames, famages, famheight)

19
Converting data

• To change a matrix to a data frame
• as.data.frame()
• To change a data frame to a matrix use
• fammatlt-as.matrix(fam)
• Try this on your own
• How does this change?
• What happens when we type fam,23 and
  fammat,23

20
Opening data

• You must always know the directory where your
  files have been stored
• In R for Windows, use \\
• c\\splus\\free.dat
• In Unix, splus/free.dat
• For now we will show to do this in Windows

21
Reading in data

• Change the directory to g\shared\bio271summer
• Lets look at the class data set in the notepad
• To read in data, use this command
  classlt-read.table("class.dat", headerT)
• This command assumes that the data is space or
  tab delimited
• Reads in as data frame

22
Working with the data

• Type class to look at the data
• How could we find people height in cm?
• Use cmheightlt-class,32.54
• This command completes the operation on the
  entire column as we discussed before
• We can attach this new variable to the old
  dataset using newclasslt-cbind(class,cmheight)
• rbind is used to combine extra rows
• In this example, that would be more students

23
Reading in data cont.

• Look at the data in auto.dat in the notepad
• Note that data is comma delimited
• Try the read.table method
• What is wrong? Look at ?read.table
• autolt-read.table(auto2.dat", sep",", headerT)

24
Practice

• Find the minimum height in the class
• Add two family members to your class data set
• Make a new variable mpg from the third column of
  auto

25
Outputting data-CHECK

• write a vector to a file write()
• write(x,fileoutdata)
• write a matrices or data frames
• write.matrixlt-function(x,file,sep)
• xlt-as.matrix(x)
• plt-ncol(x)
• cat(dimnames(x)2,format(t(x)),filefile,
• sepc(rep(sep,p-1),\n))
• Try to write your family dataset to your P\\
  drive

26
Input/Output

• execute commands from a file
• source(command.s)
• use options(echoT) to have the commands echoed.
• divert output to a file
• sink(record.lis)
• write objects to an external file
• dump(c(a,x,ink),fileoutdat)
• general print
• cat(format(iris,1,1),fill60)

27
Sorting data

• There are several ways to sort data in R
• To sort a vector, use sort()
• To sort the ages in the class, sort(class,2)
• To sort the entire matrix, use order()
• To order the class by age, classorder(class,2),
  
• To get the same result in two steps
• olt-order(class,2)
• classo,
• Try to sort auto by foreign

28
Missing values

• What happens if there are missing values as in
  the auto data sets?
• R codes these as NA
• How can we change the NAs to 0s?
• is.na(x) is a logical function that assigns a T
  to all values that are NA and F otherwise
• Ans datais.na(data)lt-0

29
Practice

• In slide 11, several functions were mentioned
  including sum and min
• Try these functions on mpg
• What has happened?
• How can we find the sum of mpg not including the
  missing values?
• mean(mpg!is.na(mpg))

30
Practice

• What is the difference when you use the functions
  sqrt on the same data?
• Why is there a difference in the effect of the
  missing data?

31
Loops and conditionals

• Conditional
• if (expr) expr
• if (expr) expr else expr
• Iteration
• repeat expr
• while (expr) expr
• for (name in expr1) expr
• For comparisons use
• for equal
• ! for not equal
• gt for greater than
• for and
• for or

32
Examples

• What happens with the following code?
• if (value1) checklt-1 else checklt-0
• counterlt-0
• for (i in 110)
• if (autoi,3lt10)counterlt-counter1

33
Basic R functions

• Lets use the class data set
• Find the summary statistics of the data
• summary(class)
• Notice how names are handled compared to ages and
  heights
• What happens when we type range(class)
• How can we find the range of age and height using
  this command?

34
Functions

• You can define a function to complete any
  operation
• outlt-function(var)definition
• Lets look at this function
• filter lt- function(x)
• if (is.na(sum(x)))fillt-F
• else fillt-T
• filt lt- apply(auto, 1, filter)
• newautolt-autofilt,
• What is this function doing?

35
Practice

• Write a function to calculate the sum of the
  numbers in a vector greater ten and the sum of
  the numbers less than or equal to ten
• Output the answer in a list using list(ans1,ans2)
  at the end of your function
• Try your function on the mpg data from the auto
  data set
• Look at the output when you apply your function

36
Generating data from distributions

• Many applications require generating data from
  specific distributions
• rltdistnamegt(n,ltparametersgt)
• Possible distributions beta, cauchy, chisq, f,
  gamma, norm, t, unif
• You find other characteristics of distributions
  as well
• dltdistgt(x,ltparametersgt) density at x
• pltdistgt(x,ltparametersgt) cumulative distribution
  function to x
• qltdistgt(p,ltparametersgt) inverse cdf

37
Using the distribution functions

• Often when we use simulations we need to use the
  r functions
• Try samplelt-runif(n10,min0,max1)
• What happens here?

38
Plots

• One of the biggest advantages of R is the quality
  of the plots
• Lets plot the ages of the class
• To make plots in R, use the following commands
  for the appropriate plots
• agelt-class,2
• histogram- hist(age)
• box plot- boxplot(age)

39
Plot Command

• The basic command-line command for producing a
  scatter plot or line graph.
• col set colors,
• lty set line types,
• lwd set line widths,
• pch set the character type,
• type pick points (type "p"), lines ("l"),
• cex set the "character expansion,
• xlab and ylab set the labels,
• xlim and ylim set the limits of the axes,
• main put a title on the plot,
• mtext add a sub-title,
• help (par) for details

40
One-Dimensional Plots

• barplot(height) simple form
• barplot(height, width, names, space.2,
  insideTRUE, besideFALSE, horizFALSE, legend,
  angle, density, col, blocksTRUE)
• boxplot(..., range, width, varwidthFALSE,
  notchFALSE, names, plotTRUE)
• hist(x, nclass, breaks, plotTRUE, angle,
  density, col, inside)

41
Two-Dimensional Plots

• lines(x, y, type"l")
• points(x, y, type"p"))
• matplot(x, y, type"p", lty15, pch, col14)
• matpoints(x, y, type"p", lty15, pch, col14)
  
• matlines(x, y, type"l", lty15, pch, col14)
• plot(x, y, type"p", log"")
• abline(coef), abline(a, b), abline(reg),
  abline(h), abline(v)
• qqplot(x, y, plotTRUE)
• qqnorm(x, dataxFALSE, plotTRUE)

42
Three-Dimensional Plots

• contour(x, y, z, v, nint5, addFALSE, labex)
• interp(x, y, z, xo, yo, ncp0, extrapFALSE)
• persp(z, eyec(-6,-8,5), ar1)

43
Multiple Plots Per Page

• par(mfrowc(nrow, ncol), omac(0, 0, 4, 0))
• mfrowc(m,n) subsequent figures will be drawn
  row-by-row in an m by n matrix on the page.
• omac(xbot,xlef,xtop,xrig)outer margin lines of
  text.
• mtext(side3, line0, cex2, outerT, "This is an
  Overall Title For the Page")
• Try this code on your own
• par(mfrowc(2,1)
• hist(age)
• plot(class,2,class,3)

44
Output to a postscript file

• Often we want to output an R graph to a
  postscript file to place it into a Latex file or
  other document
• To do this, we use the following code
• postscript(graph1.ps) This opens a postscript
  file in the home directory
• plot(regr) This plots a graph into the file
• dev.off() This closes the postscript file

45
Making plots of your own

• Make the following plots
• Histogram of height in the class with the
  appropriate labels
• Scatterplot of height and age in the class using
  a different point
• Make a postscript file with four plots of your
  choice from the baby dataset on one graph
• Write a function to make a histogram and boxplot
  on one graph and use it on any of your data