An Introduction to the R Statistical Programming Language

1
An Introduction to the RStatistical Programming
Language

• Timothy J. Robinson
• University of Wyoming
• tjrobin_at_uwyo.edu
• and
• Doug Nychka
• (add info)

Some material from NY Times, various Harry Potter
websites, Matthew Keller (www.mckeller.com),
Revolution Analytics, and Ian Westbrooke (NZDOC)
2
What is R?

• A. 18th letter of the alphabet
• B. Rating for movies that suggests our kids
  should refrain from watching
• C. Favorite pirate saying
• Fastest growing statistical software package in
  the world
• All of the above

3
Users of R

• Google, Pfizer, Merck, Bank of America, Procter
  Gamble, GE, Shell, Intercontinental Hotels, IBM,
  Hewlett-Packert, and many others
• National laboratories (Los Alamos, Sandia, and
  others), U.S. Fish and Wildlife Service,
  U.S.G.S., U.S. Department of Treasury, and many
  others
• Growing number of academic institutions in the
  U.S. and around the world
• In January 2009, estimated that over 2 million
  users of R worldwide

4
Where did it come from?

• First appeared in 1996 when professors Ross Ihaka
  and Robert Gentleman, University of Auckland in
  New Zealand released the code as a free software
  package
• John Chambers, Bell Labs Researcher and currently
  a consultant for the Department of Statistics at
  Stanford, was an early champion of R. Chambers
  was part of the development team for SPLUS
• Recently, Revolution Analytics has developed a
  platform for enabling R to handle massive data
  sets
• NY Times 2009 suggested that R is rivaling the
  biggest stat software company of all, SASthe
  response of SAS in the article

5
(No Transcript)
6
Loading R

• Go to http//cran.r-project/
• R for Windows screen, double-click base
  
• Find, click on download R save the executable
  file
• Double-click the executable file to get into the
  installation wizard
• When finished, you will have the R-icon for the
  particular version on your desktop

7
(No Transcript)
8
At http//cran.r-project.org/
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
After Double-Clicking on R icon
15
The R Prompt (gt)

• gt This is the R prompt. It says R is
  ready to take your command
• R as a calculator
• gt 52
• gt 62
• gt sqrt(16)
• Creating a data set
• gt flowers lt- c(1.2, 2.6, 3.4, 2.9,5.1)
• gt flowers
• The above line tells R to print the data set
  flowers

Red denotes commands Blue denotes output
16
Important Rules

• We created a variable
• Variable names are case sensitive
• No blanks in name
• (can use _ or . to join words, but better not to
  use a blank between names)
• Start with a letter (cap or lc)
• Use of lt- instead of

17
Non-Numeric Data

• Enclose in quotes, single or double
• Separate entries with comma
• Example
• gt names c(Tim, Dawn, Ian, Rosamund,
  Jay)

18
Common Statistics Functions in R

• function calculates this
• mean ( ) mean
• median ( ) median
• mode ( ) mode
• sd ( ) standard deviation
• range ( ) range
• IQR ( ) interquartile range
• min ( ) minimum value
• max ( ) maximum value
• cor ( ) correlation
• quantile ( ) percentile
• length( ) sample size

19
R programming language is a lot like magic...
except instead of spells you have functions.
20
Fathom, MINITAB, SAS, etc. users are like
muggles - cant change their environment
- rely on algorithms which are canned
- rely on the output provided - have
to pay to constraining algorithms.
21

wizard
R users are like wizards. They can rely on
functions (spells) that have been developed for
them by statistical researchers, but they can
also create their own. They dont have to pay for
the use of them, and once experienced enough
(like Dumbledore), they are almost unlimited in
their ability to change their environment.
Matthew Keller
22
What Drives the Capability of R?

• Installed packages/libraries
• The base installation includes many
  packages/libraries but you can supplement the
  base installation with other packages/libraries
  

23
Just scroll down to one of the many USA sites
24
Learning R....
25
R-help listserve....
26
Rcmdr

• It is a package in R
• Advantages
• Click on the menu similar to SPSS, Minitab, JMP
• No need to memorize the R-functions
• Save the script in your personnel directory and
  use it for future.
• Good for non-statistical people or folks just
  starting use of R
• Disadvantages
• No access for some functions
• Limited access to the packages
• Does not always help those with limited
  statistical skills enough

27
Installing Rcmdr
Choose the Rcmdr package and click OK
28
Installing vs. Loading a Package/Library

• Installing a package involves getting your
  machine to talk to a CRAN server and downloading
  the package/library from the CRAN site to your
  personal machine
• Once the package is installed, we must load the
  package in order to activate it for our current
  sessionthis can be accomplished in one of two
  ways - Use the library( )
  function - Use the drop-down menu from the
  console window

29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
You now have a point and click
environment within which to operate