data.frame objects in R (via “R in Action”)

The followings introductory post is intended for new users of R.  It deals with R data frames: what they are, and how to create, view, and update them. This is a guest article by Dr. Robert I. Kabacoff, the founder of (one of) the first online R tutorials websites: Quick-R.  Kabacoff has recently published the book ”R […]

The followings introductory post is intended for new users of R.  It deals with R data frames: what they are, and how to create, view, and update them.

This is a guest article by Dr. Robert I. Kabacoff, the founder of (one of) the first online R tutorials websites: Quick-R.  Kabacoff has recently published the book ”R in Action“, providing a detailed walk-through for the R language based on various examples for illustrating R’s features (data manipulation, statistical methods, graphics, and so on…)

For readers of this blog, there is a 38% discount off the “R in Action” book (as well as all other eBooks, pBooks and MEAPs at Manning publishing house), simply by using the code rblogg38 when reaching checkout.

Let us now talk about data frames:

Data Frames


A data frame is more general than a matrix in that different columns can contain different modes of data (numeric, character, and so on). It’s similar to the datasets you’d typically see in SAS, SPSS, and Stata. Data frames are the most common data structure you’ll deal with in R.

The patient dataset in table 1 consists of numeric and character data.

Table 1: A patient dataset

PatientID

AdmDate

Age

Diabetes

Status

110/15/200925Type1Poor
211/01/200934Type2Improved
310/21/200928Type1Excellent
410/28/200952Type1Poor

Because there are multiple modes of data, you can’t contain this data in a matrix. In this case, a data frame would be the structure of choice.

A data frame is created with the data.frame() function:

1
mydata <- data.frame(col1, col2, col3,…)

where col1, col2, col3, … are column vectors of any type (such as character, numeric, or logical). Names for each column can be provided with the names function.

The following listing makes this clear.

Listing 1 Creating a data frame

1
2
3
4
5
6
7
8
9
10
11
> patientID <- c(1, 2, 3, 4)
> age <- c(25, 34, 28, 52)
> diabetes <- c("Type1", "Type2", "Type1", "Type1")
> status <- c("Poor", "Improved", "Excellent", "Poor")
> patientdata <- data.frame(patientID, age, diabetes, status)
> patientdata
  patientID age diabetes status
1         1  25    Type1 Poor
2         2  34    Type2 Improved
3         3  28    Type1 Excellent
4         4  52    Type1 Poor

Each column must have only one mode, but you can put columns of different modes together to form the data frame. Because data frames are close to what analysts typically think of as datasets, we’ll use the terms columns and variables interchangeably when discussing data frames.

There are several ways to identify the elements of a data frame. You can use the subscript notation or you can specify column names. Using the patientdata data frame created earlier, the following listing demonstrates these approaches.

Listing 2 Specifying elements of a data frame

1
2
3
4
5
6
7
8
9
10
11
12
13
14
> patientdata[1:2]
  patientID age
1         1  25
2         2  34
3         3  28
4         4  52
> patientdata[c("diabetes", "status")]
  diabetes status
1    Type1 Poor
2    Type2 Improved
3    Type1 Excellent
4    Type1 Poor
> patientdata$age    #age variable in the patient data frame
[1] 25 34 28 52

The $ notation in the third example is used to indicate a particular variable from a given data frame. For example, if you want to cross-tabulate diabetes type by status, you could use the following code:

1
2
3
4
5
> table(patientdata$diabetes, patientdata$status)
 
        Excellent Improved Poor
  Type1         1        0    2
  Type2         0        1    0

It can get tiresome typing patientdata$ at the beginning of every variable name, so shortcuts are available. You can use either the attach() and detach() or with() functions to simplify your code.

attach, detach, and with

The attach() function adds the data frame to the R search path. When a variable name is encountered, data frames in the search path are checked in order to locate the variable. Using a sample (mtcars) data frame, you could use the following code to obtain summary statistics for automobile mileage (mpg), and plot this variable against engine displacement (disp), and weight (wt):

1
2
3
summary(mtcars$mpg)
plot(mtcars$mpg, mtcars$disp)
plot(mtcars$mpg, mtcars$wt)

This could also be written as

1
2
3
4
5
attach(mtcars)
  summary(mpg)
  plot(mpg, disp)
  plot(mpg, wt)
detach(mtcars)

The detach() function removes the data frame from the search path. Note that detach() does nothing to the data frame itself. The statement is optional but is good programming practice and should be included routinely.

The limitations with this approach are evident when more than one object can have the same name. Consider the following code:

1
2
3
4
5
6
7
8
9
> mpg <- c(25, 36, 47)
> attach(mtcars)
 
The following object(s) are masked _by_ ‘.GlobalEnv: mpg
> plot(mpg, wt)
Error in xy.coords(x, y, xlabel, ylabel, log) :
  ‘x’ and ‘y’ lengths differ
> mpg
[1] 25 36 47

Here we already have an object named mpg in our environment when the mtcars data frame is attached. In such cases, the original object takes precedence, which isn’t what you want. The plot statement fails because mpg has 3 elements and disp has 32 elements. The attach() and detach() functions are best used when you’re analyzing a single data frame and you’re unlikely to have multiple objects with the same name. In any case, be vigilant for warnings that say that objects are being masked.

An alternative approach is to use the with() function. You could write the previous example as

1
2
3
4
5
with(mtcars, {
  summary(mpg, disp, wt)
  plot(mpg, disp)
  plot(mpg, wt)
})

In this case, the statements within the {} brackets are evaluated with reference to the mtcars data frame. You don’t have to worry about name conflicts here. If there’s only one statement (for example, summary(mpg)), the {} brackets are optional.

The limitation of the with() function is that assignments will only exist within the function brackets. Consider the following:

1
2
3
4
5
6
7
8
> with(mtcars, {
   stats <- summary(mpg)
   stats
  })
   Min. 1st Qu. Median Mean 3rd Qu. Max.
  10.40 15.43 19.20 20.09 22.80 33.90
> stats
Error: object ‘stats’ not found

If you need to create objects that will exist outside of the with() construct, use the special assignment operator <<- instead of the standard one (<-). It will save the object to the global environment outside of the with() call. This can be demonstrated with the following code:

1
2
3
4
5
6
7
8
9
> with(mtcars, {
   nokeepstats <- summary(mpg)
   keepstats <<- summary(mpg)
})
> nokeepstats
Error: object ‘nokeepstats’ not found
> keepstats
   Min. 1st Qu. Median Mean 3rd Qu. Max.
    10.40 15.43 19.20 20.09 22.80 33.90

Most books on R recommend using with() over attach(). I think that ultimately the choice is a matter of preference and should be based on what you’re trying to achieve and your understanding of the implications.

Case identifiers

In the patient data example, patientID is used to identify individuals in the dataset. In R, case identifiers can be specified with a rowname option in the data frame function. For example, the statement

1
2
patientdata <- data.frame(patientID, age, diabetes, status,
   row.names=patientID)

specifies patientID as the variable to use in labeling cases on various printouts and graphs produced by R.

Summary

One of the most challenging tasks in data analysis is data preparation. R provides various structures for holding data and many methods for importing data from both keyboard and external sources. One of those structures is data frames, which we covered here. Your ability to specify elements of these structures via the bracket notation is particularly important in selecting, subsetting, and transforming data.

R offers a wealth of functions for accessing external data. This includes data from flat files, web files, statistical packages, spreadsheets, and databases. Note that you can also export data from R into these external formats. We showed you how to use either the attach() and detach() or with() functions to simplify your code.

This article first appeared as chapter 2.2.4 from the “R in action book, and is published with permission from Manning publishing house.

Book review: 25 Recipes for Getting Started with R

Recently I was asked by O’Reilly publishing to give a book review for Paul Teetor new introductory book to R.  After giving the book some attention and appreciating it’s delivery of the material, I was happy to write and post this review.  Also, I’m very happy to see how a major publishing house like O’Reilly is producing more and more R books, great news indeed.

And now for the book review:

Executive summary: a book that offers a well designed gentle introduction for people with some background in statistics wishing to learn how to get common (basic) tasks done with R.

Information

By: Paul Teetor
Publisher:O’Reilly
MediaReleased: January 2011
Pages: 58 (est.)

Format

The book “25 Recipes for Getting Started with R” offers an interesting take on how to bring R to the general (statistically oriented) public.

Continue reading “Book review: 25 Recipes for Getting Started with R”

Tips for the R beginner (a 5 page overview)

In this post I publish a PDF document titled “A collection of tips for R in Finance”.
It is a basic 5 page introduction to R in finances by Arnaud Amsellem (linked in profile).

The article offers tips related to the following points:

  • Code Editor
  • Organizing R code
  • Update packages
  • Getting external data into R
  • Communicating with external applications
  • Optimizing R code

This article is well articulated, and offers a perspective of someone who is experienced in the field and touches points that I can imagine beginners might otherwise overlook. I hope publishing it here will be of use to some readers out there.

Update: as some readers have noted to me (by e-mail, and by commenting), this document touches very lightly on the topic of “finances” in R. I therefore decided to update the title from “R in finance – some tips for beginners”, to it’s current form.

Lastly: if you (a reader of this blog) feel you have an article (“post”) to contribute, but don’t feel like starting your own blog, feel welcome to contact me, and I’ll be glad to post what you have to say on my blog (and subsequently, also on R bloggers).

Here is the article:
Continue reading “Tips for the R beginner (a 5 page overview)”

Rose plot using Deducers ggplot2 plot builder

The (excellent!) LearnR blog had a post today about making a rose plot in
ggplot2.

Following today’s announcement, by Ian Fellows, regarding the release of the new version of Deducer (0.4) offering a strong support for ggplot2 using a GUI plot builder, Ian also sent an e-mail where he shows how to create a rose plot using the new ggplot2 GUI included in the latest version of Deducer. After the template is made, the plot can be generated with 4 clicks of the mouse.

Here is a video tutorial (Ian published) to show how this can be used:

The generated template file is available at:
http://neolab.stat.ucla.edu/cranstats/rose.ggtmpl

I am excited about the work Ian is doing, and hope to see more people publish use cases with Deducer.

ggplot2 plot builder is now on CRAN! (through Deducer 0.4 GUI for R)

Ian fellows, a hard working contributer to the R community (and a cool guy), has announced today the release of Deducer (0.4) to CRAN (scheduled to update in the next day or so).
This major update also includes the release of a new plug-in package (DeducerExtras), containing additional dialogs and functionality.

Following is the e-mail he sent out with all the details and demo videos.

Continue reading “ggplot2 plot builder is now on CRAN! (through Deducer 0.4 GUI for R)”

Repeated measures ANOVA with R (functions and tutorials)

Repeated measures ANOVA is a common task for the data analyst.

There are (at least) two ways of performing “repeated measures ANOVA” using R but none is really trivial, and each way has it’s own complication/pitfalls (explanation/solution to which I was usually able to find through searching in the R-help mailing list).

So for future reference, I am starting this page to document links I find to tutorials, explanations (and troubleshooting) of “repeated measure ANOVA” done with R

Functions and packages

(I suggest using the tutorials supplied bellow for how to use these functions)

  • aov {stats} – offers SS type I repeated measures anova, by a call to lm for each stratum. A short example is given in the ?aov help file
  • Anova {car} – Calculates type-II or type-III analysis-of-variance tables for model objects produced by lm, and for various other object. The ?Anova help file offers an example for how to use this for repeated measures
  • ezANOVA {ez} – This function provides easy analysis of data from factorial experiments, including purely within-Ss designs (a.k.a. “repeated measures”), purely between-Ss designs, and mixed within-and-between-Ss designs, yielding ANOVA results and assumption checks. It is a wrapper of the Anova {car} function, and is easier to use. The ez package also offers the functions ezPlot and ezStats to give plot and statistics of the ANOVA analysis. The ?ezANOVA help file gives a good demonstration for the functions use (My thanks goes to Matthew Finkbe for letting me know about this cool package)
  • friedman.test {stats} – Performs a Friedman rank sum test with unreplicated blocked data. That is, a non-parametric one-way repeated measures anova. I also wrote a wrapper function to perform and plot a post-hoc analysis on the friedman test results
  • Non parametric multi way repeated measures anova – I believe such a function could be developed based on the Proportional Odds Model, maybe using the {repolr} or the {ordinal} packages. But I still didn’t come across any function that implements these models (if you do – please let me know in the comments).
  • Repeated measures, non-parametric, multivariate analysis of variance – as far as I know, such a method is not currently available in R.  There is, however, the Analysis of similarities (ANOSIM) analysis which provides a way to test statistically whether there is a significantdifference between two or more groups of sampling units.  Is is available in the {vegan} package through the “anosim” function.  There is also a tutorial and a relevant published paper.

Good Tutorials

Troubelshooting

Unbalanced design
Unbalanced design doesn’t work when doing repeated measures ANOVA with aov, it just doesn’t. This situation occurs if there are missing values in the data or that the data is not from a fully balanced design. The way this will show up in your output is that you will see the between subject section showing withing subject variables.

A solution for this might be to use the Anova function from library car with parameter type=”III”. But before doing that, first make sure you understand the difference between SS type I, II and III. Here is a good tutorial for helping you out with that.
By the way, these links are also useful in case you want to do a simple two way ANOVA for unbalanced design

I will “later” add R-help mailing list discussions that I found helpful on the subject.

If you come across good resources, please let me know about them in the comments.

Correlation scatter-plot matrix for ordered-categorical data

When analyzing a questionnaire, one often wants to view the correlation between two or more Likert questionnaire item’s (for example: two ordered categorical vectors ranging from 1 to 5).

When dealing with several such Likert variable’s, a clear presentation of all the pairwise relation’s between our variable can be achieved by inspecting the (Spearman) correlation matrix (easily achieved in R by using the “cor.test” command on a matrix of variables).
Yet, a challenge appears once we wish to plot this correlation matrix. The challenge stems from the fact that the classic presentation for a correlation matrix is a scatter plot matrix – but scatter plots don’t (usually) work well for ordered categorical vectors since the dots on the scatter plot often overlap each other.

There are four solution for the point-overlap problem that I know of:

  1. Jitter the data a bit to give a sense of the “density” of the points
  2. Use a color spectrum to represent when a point actually represent “many points”
  3. Use different points sizes to represent when there are “many points” in the location of that point
  4. Add a LOWESS (or LOESS) line to the scatter plot – to show the trend of the data

In this post I will offer the code for the  a solution that uses solution 3-4 (and possibly 2, please read this post comments). Here is the output (click to see a larger image):

And here is the code to produce this plot:

Continue reading “Correlation scatter-plot matrix for ordered-categorical data”

A nice link: "Some hints for the R beginner"

Patrick Burns just posted to the mailing list the following massage:

There is now a document called “Some hints for the R beginner” whose purpose is to get people up and running with R as quickly as possible.

Direct access to it is:
http://www.burns-stat.com/pages/Tutor/hints_R_begin.html

JRR Tolkien wrote a story (sans hobbits) called ‘Leaf by Niggle’ that has always resonated with me. I offer you an imperfect, incomplete tree (but my roof is intact).

Suggestions for improvements are encouraged.

And here is the link tree for the document (for your easy reviewing of the offered content) :

Continue reading “A nice link: "Some hints for the R beginner"”

Web Development with R – an HD video tutorial of Jeroen Ooms talk

Here is a HD version of a video tutorial on web development with R, a lecture that was given by Jeroen Ooms (the guy who made A web application for R’s ggplot2). This talk was given at the Bay Area UseR Group meeting on R-Powered Web Apps.

You can also view the slides for his talk and view (great) examples for: stockplotlme4, and gpplot2.

Thanks again to Jeroen for sharing his knowledge and experience!