About Me Blog
A tutorial on tidy cross-validation with R Analyzing NetHack data, part 1: What kills the players Analyzing NetHack data, part 2: What players kill the most Building a shiny app to explore historical newspapers: a step-by-step guide Classification of historical newspapers content: a tutorial combining R, bash and Vowpal Wabbit, part 1 Classification of historical newspapers content: a tutorial combining R, bash and Vowpal Wabbit, part 2 Curly-Curly, the successor of Bang-Bang Dealing with heteroskedasticity; regression with robust standard errors using R Easy time-series prediction with R: a tutorial with air traffic data from Lux Airport Exporting editable plots from R to Powerpoint: making ggplot2 purrr with officer Fast food, causality and R packages, part 1 Fast food, causality and R packages, part 2 For posterity: install {xml2} on GNU/Linux distros Forecasting my weight with R From webscraping data to releasing it as an R package to share with the world: a full tutorial with data from NetHack Get text from pdfs or images using OCR: a tutorial with {tesseract} and {magick} Getting data from pdfs using the pdftools package Getting the data from the Luxembourguish elections out of Excel Going from a human readable Excel file to a machine-readable csv with {tidyxl} Historical newspaper scraping with {tesseract} and R How Luxembourguish residents spend their time: a small {flexdashboard} demo using the Time use survey data Imputing missing values in parallel using {furrr} Intermittent demand, Croston and Die Hard Looking into 19th century ads from a Luxembourguish newspaper with R Making sense of the METS and ALTO XML standards Manipulate dates easily with {lubridate} Manipulating strings with the {stringr} package Maps with pie charts on top of each administrative division: an example with Luxembourg's elections data Missing data imputation and instrumental variables regression: the tidy approach Modern R with the tidyverse is available on Leanpub Objects types and some useful R functions for beginners Pivoting data frames just got easier thanks to `pivot_wide()` and `pivot_long()` R or Python? Why not both? Using Anaconda Python within R with {reticulate} Searching for the optimal hyper-parameters of an ARIMA model in parallel: the tidy gridsearch approach Some fun with {gganimate} Split-apply-combine for Maximum Likelihood Estimation of a linear model Statistical matching, or when one single data source is not enough The best way to visit Luxembourguish castles is doing data science + combinatorial optimization The never-ending editor war (?) The year of the GNU+Linux desktop is upon us: using user ratings of Steam Play compatibility to play around with regex and the tidyverse Using Data Science to read 10 years of Luxembourguish newspapers from the 19th century Using a genetic algorithm for the hyperparameter optimization of a SARIMA model Using cosine similarity to find matching documents: a tutorial using Seneca's letters to his friend Lucilius Using linear models with binary dependent variables, a simulation study Using the tidyverse for more than data manipulation: estimating pi with Monte Carlo methods What hyper-parameters are, and what to do with them; an illustration with ridge regression {disk.frame} is epic {pmice}, an experimental package for missing data imputation in parallel using {mice} and {furrr} Building formulae Functional peace of mind Get basic summary statistics for all the variables in a data frame Getting {sparklyr}, {h2o}, {rsparkling} to work together and some fun with bash Importing 30GB of data into R with sparklyr Introducing brotools It's lists all the way down It's lists all the way down, part 2: We need to go deeper Keep trying that api call with purrr::possibly() Lesser known dplyr 0.7* tricks Lesser known dplyr tricks Lesser known purrr tricks Make ggplot2 purrr Mapping a list of functions to a list of datasets with a list of columns as arguments Predicting job search by training a random forest on an unbalanced dataset Teaching the tidyverse to beginners Why I find tidyeval useful tidyr::spread() and dplyr::rename_at() in action Easy peasy STATA-like marginal effects with R Functional programming and unit testing for data munging with R available on Leanpub How to use jailbreakr My free book has a cover! Work on lists of datasets instead of individual datasets by using functional programming Method of Simulated Moments with R New website! Nonlinear Gmm with R - Example with a logistic regression Simulated Maximum Likelihood with R Bootstrapping standard errors for difference-in-differences estimation with R Careful with tryCatch Data frame columns as arguments to dplyr functions Export R output to a file I've started writing a 'book': Functional programming and unit testing for data munging with R Introduction to programming econometrics with R Merge a list of datasets together Object Oriented Programming with R: An example with a Cournot duopoly R, R with Atlas, R with OpenBLAS and Revolution R Open: which is fastest? Read a lot of datasets at once with R Unit testing with R Update to Introduction to programming econometrics with R Using R as a Computer Algebra System with Ryacas

Objects types and some useful R functions for beginners

This blog post is an excerpt of my ebook Modern R with the tidyverse that you can read for free here. This is taken from Chapter 2, which explains the different R objects you can manipulate as well as some functions to get you started.

Objects, types and useful R functions to get started

All objects in R have a given type. You already know most of them, as these types are also used in mathematics. Integers, floating point numbers, or floats, matrices, etc, are all objects you are already familiar with. But R has other, maybe lesser known data types (that you can find in a lot of other programming languages) that you need to become familiar with. But first, we need to learn how to assign a value to a variable. This can be done in two ways:

a <- 3


a = 3

in very practical terms, there is no difference between the two. I prefer using <- for assigning values to variables and reserve = for passing arguments to functions, for example:

spam <- mean(x = c(1,2,3))

I think this is less confusing than:

spam = mean(x = c(1,2,3))

but as I explained above you can use whatever you feel most comfortable with.

The numeric class

To define single numbers, you can do the following:

a <- 3

The class() function allows you to check the class of an object:

## [1] "numeric"

Decimals are defined with the character .:

a <- 3.14

R also supports integers. If you find yourself in a situation where you explicitly need an integer and not a floating point number, you can use the following:

a  <- as.integer(3)
## [1] "integer"

The as.integer() function is very useful, because it converts its argument into an integer. There is a whole family of as.*() functions. To convert a into a floating point number again:

## [1] "numeric"

There is also is.numeric() which tests whether a number is of the numeric class:

## [1] TRUE

These functions are very useful, there is one for any of the supported types in R. Later, we are going to learn about the {purrr} package, which is a very powerful package for functional programming. This package includes further such functions.

The character class

Use " " to define characters (called strings in other programming languages):

a <- "this is a string"
## [1] "character"

To convert something to a character you can use the as.character() function:

a <- 4.392

## [1] "numeric"
## [1] "character"

It is also possible to convert a character to a numeric:

a <- "4.392"

## [1] "character"
## [1] "numeric"

But this only works if it makes sense:

a <- "this won't work, chief"

## [1] "character"
## Warning: NAs introduced by coercion
## [1] NA

A very nice package to work with characters is {stringr}, which is also part of the {tidyverse}.

The factor class

Factors look like characters, but are very different. They are the representation of categorical variables. A {tidyverse} package to work with factors is {forcats}. You would rarely use factor variables outside of datasets, so for now, it is enough to know that this class exists. We are going to learn more about factor variables in Chapter 4, by using the {forcats} package.

The Date class

Dates also look like characters, but are very different too:

## [1] "2019-03-19"
## [1] "Date"

Manipulating dates and time can be tricky, but thankfully there’s a {tidyverse} package for that, called {lubridate}. We are going to go over this package in Chapter 4.

The logical class

This class is the result of logical comparisons, for example, if you type:

4 > 3
## [1] TRUE

R returns TRUE, which is an object of class logical:

k <- 4 > 3
## [1] "logical"

In other programming languages, logicals are often called bools. A logical variable can only have two values, either TRUE or FALSE. You can test the truthiness of a variable with isTRUE():

k <- 4 > 3
## [1] TRUE

How can you test if a variable is false? There is not a isFALSE() function (at least not without having to load a package containing this function), but there is way to do it:

k <- 4 > 3
## [1] FALSE

The ! operator indicates negation, so the above expression could be translated as is k not TRUE?. There are other such operators, namely &, &&, |, ||. & means and and | stands for or. You might be wondering what the difference between & and && is? Or between | and ||? & and | work on vectors, doing pairwise comparisons:

one & two

Compare this to the && operator:

one && two
## [1] FALSE

The && and || operators only compare the first element of the vectors and stop as soon as a the return value can be safely determined. This is called short-circuiting. Consider the following:

three <- c(TRUE, TRUE, FALSE, FALSE)
one && two && three
## [1] FALSE
one || two || three
## [1] TRUE

The || operator stops as soon it evaluates to TRUE whereas the && stops as soon as it evaluates to FALSE. Personally, I rarely use || or && because I get confused. I find using | or & in combination with the all() or any() functions much more useful:

any(one & two)
## [1] TRUE
all(one & two)
## [1] FALSE

any() checks whether any of the vector’s elements are TRUE and all() checks if all elements of the vector are TRUE.

As a final note, you should know that is possible to use T for TRUE and F for FALSE but I would advise against doing this, because it is not very explicit.

Vectors and matrices

You can create a vector in different ways. But first of all, it is important to understand that a vector in most programming languages is nothing more than a list of things. These things can be numbers (either integers or floats), strings, or even other vectors. A vector in R can only contain elements of one single type. This is not the case for a list, which is much more flexible. We will talk about lists shortly, but let’s first focus on vectors and matrices.

The c() function

A very important function that allows you to build a vector is c():

a <- c(1,2,3,4,5)

This creates a vector with elements 1, 2, 3, 4, 5. If you check its class:

## [1] "numeric"

This can be confusing: you where probably expecting a to be of class vector or something similar. This is not the case if you use c() to create the vector, because c() doesn’t build a vector in the mathematical sense, but a so-called atomic vector. Checking its dimension:


returns NULL because an atomic vector doesn’t have a dimension. If you want to create a true vector, you need to use cbind() or rbind().

But before continuing, be aware that atomic vectors can only contain elements of the same type:

c(1, 2, "3")
## [1] "1" "2" "3"

because “3” is a character, all the other values get implicitly converted to characters. You have to be very careful about this, and if you use atomic vectors in your programming, you have to make absolutely sure that no characters or logicals or whatever else are going to convert your atomic vector to something you were not expecting.

cbind() and rbind()

You can create a true vector with cbind():

a <- cbind(1, 2, 3, 4, 5)

Check its class now:

## [1] "matrix"

This is exactly what we expected. Let’s check its dimension:

## [1] 1 5

This returns the dimension of a using the LICO notation (number of LInes first, the number of COlumns).

It is also possible to bind vectors together to create a matrix.

b <- cbind(6,7,8,9,10)

Now let’s put vector a and b into a matrix called matrix_c using rbind(). rbind() functions the same way as cbind() but glues the vectors together by rows and not by columns.

matrix_c <- rbind(a,b)
##      [,1] [,2] [,3] [,4] [,5]
## [1,]    1    2    3    4    5
## [2,]    6    7    8    9   10

The matrix class

R also has support for matrices. For example, you can create a matrix of dimension (5,5) filled with 0’s with the matrix() function:

matrix_a <- matrix(0, nrow = 5, ncol = 5)

If you want to create the following matrix:

\[ B = \left( \begin{array}{ccc} 2 & 4 & 3 \\ 1 & 5 & 7 \end{array} \right) \]

you would do it like this:

B <- matrix(c(2, 4, 3, 1, 5, 7), nrow = 2, byrow = TRUE)

The option byrow = TRUE means that the rows of the matrix will be filled first.

You can access individual elements of matrix_a like so:

matrix_a[2, 3]
## [1] 0

and R returns its value, 0. We can assign a new value to this element if we want. Try:

matrix_a[2, 3] <- 7

and now take a look at matrix_a again.

##      [,1] [,2] [,3] [,4] [,5]
## [1,]    0    0    0    0    0
## [2,]    0    0    7    0    0
## [3,]    0    0    0    0    0
## [4,]    0    0    0    0    0
## [5,]    0    0    0    0    0

Recall our vector b:

b <- cbind(6,7,8,9,10)

To access its third element, you can simply write:

## [1] 8

I have heard many people praising R for being a matrix based language. Matrices are indeed useful, and statisticians are used to working with them. However, I very rarely use matrices in my day to day work, and prefer an approach based on data frames (which will be discussed below). This is because working with data frames makes it easier to use R’s advanced functional programming language capabilities, and this is where R really shines in my opinion. Working with matrices almost automatically implies using loops and all the iterative programming techniques, à la Fortran, which I personally believe are ill-suited for interactive statistical programming (as discussed in the introduction).

The list class

The list class is a very flexible class, and thus, very useful. You can put anything inside a list, such as numbers:

list1 <- list(3, 2)

or other lists constructed with c():

list2 <- list(c(1, 2), c(3, 4))

you can also put objects of different classes in the same list:

list3 <- list(3, c(1, 2), "lists are amazing!")

and of course create list of lists:

my_lists <- list(list1, list2, list3)

To check the contents of a list, you can use the structure function str():

## List of 3
##  $ :List of 2
##   ..$ : num 3
##   ..$ : num 2
##  $ :List of 2
##   ..$ : num [1:2] 1 2
##   ..$ : num [1:2] 3 4
##  $ :List of 3
##   ..$ : num 3
##   ..$ : num [1:2] 1 2
##   ..$ : chr "lists are amazing!"

or you can use RStudio’s Environment pane:

You can also create named lists:

list4 <- list("a" = 2, "b" = 8, "c" = "this is a named list")

and you can access the elements in two ways:

## [1] 2

or, for named lists:

## [1] "this is a named list"

Lists are used extensively because they are so flexible. You can build lists of datasets and apply functions to all the datasets at once, build lists of models, lists of plots, etc… In the later chapters we are going to learn all about them. Lists are central objects in a functional programming workflow for interactive statistical analysis.

The data.frame and tibble classes

In the next chapter we are going to learn how to import datasets into R. Once you import data, the resulting object is either a data.frame or a tibble depending on which package you used to import the data. tibbles extend data.frames so if you know about data.frame objects already, working with tibbles will be very easy. tibbles have a better print() method, and some other niceties.

However, I want to stress that these objects are central to R and are thus very important; they are actually special cases of lists, discussed above. There are different ways to print a data.frame or a tibble if you wish to inspect it. You can use View(my_data) to show the my_data data.frame in the View pane of RStudio:

You can also use the str() function:


And if you need to access an individual column, you can use the $ sign, same as for a list:



We will learn more about formulas later, but because it is an important object, it is useful if you already know about them early on. A formula is defined in the following way:

my_formula <- ~x

## [1] "formula"

Formula objects are defined using the ~ symbol. Formulas are useful to define statistical models, for example for a linear regression:

lm(y ~ x)

or also to define anonymous functions, but more on this later.


A statistical model is an object like any other in R:


my_model <- lm(mpg ~ hp, mtcars)

## [1] "lm"

my_model is an object of class lm. You can apply different functions to a model object:

## Call:
## lm(formula = mpg ~ hp, data = mtcars)
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -5.7121 -2.1122 -0.8854  1.5819  8.2360 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 30.09886    1.63392  18.421  < 2e-16 ***
## hp          -0.06823    0.01012  -6.742 1.79e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## Residual standard error: 3.863 on 30 degrees of freedom
## Multiple R-squared:  0.6024, Adjusted R-squared:  0.5892 
## F-statistic: 45.46 on 1 and 30 DF,  p-value: 1.788e-07

This class will be explored in later chapters.

NULL, NA and NaN

The NULL, NA and NaN classes are pretty special. NULL is returned when the result of function is undetermined. For example, consider list4:

## $a
## [1] 2
## $b
## [1] 8
## $c
## [1] "this is a named list"

if you try to access an element that does not exist, such as d, you will get NULL back:


NaN means “Not a Number” and is returned when a function return something that is not a number:

## Warning in sqrt(-1): NaNs produced
## [1] NaN


## [1] NaN

Basically, numbers that cannot be represented as floating point numbers are NaN.

Finally, there’s NA which is closely related to NaN but is used for missing values. NA stands for Not Available. There are several types of NAs:

  • NA_integer_
  • NA_real_
  • NA_complex_
  • NA_character_

but these are in principle only used when you need to program your own functions and need to explicitly test for the missingness of, say, a character value.

To test whether a value is NA, use the is.na() function.

Useful functions to get you started

This section will list several basic R functions that are very useful and should be part of your toolbox.


There are several functions that create sequences, seq(), seq_along() and rep(). rep() is easy enough:

rep(1, 10)
##  [1] 1 1 1 1 1 1 1 1 1 1

This simply repeats 1 10 times. You can repeat other objects too:

rep("HAHA", 10)

To create a sequence, things are not as straightforward. There is seq():

seq(1, 10)
##  [1]  1  2  3  4  5  6  7  8  9 10
seq(70, 80)
##  [1] 70 71 72 73 74 75 76 77 78 79 80

It is also possible to provide a by argument:

seq(1, 10, by = 2)
## [1] 1 3 5 7 9

seq_along() behaves similarly, but returns the length of the object passed to it. So if you pass list4 to seq_along(), it will return a sequence from 1 to 3:

## [1] 1 2 3

which is also true for seq() actually:

## [1] 1 2 3

but these two functions behave differently for arguments of length equal to 1:

##  [1]  1  2  3  4  5  6  7  8  9 10
## [1] 1

So be quite careful about that. I would advise you do not use seq(), but only seq_along() and seq_len(). seq_len() only takes arguments of length 1:

##  [1]  1  2  3  4  5  6  7  8  9 10
## [1] 1

The problem with seq() is that it is unpredictable; depending on its input, the output will either be an integer or a sequence. When programming, it is better to have function that are stricter and fail when confronted to special cases, instead of returning some result. This is a bit of a recurrent issue with R, and the functions from the {tidyverse} mitigate this issue by being stricter than their base R counterparts. For example, consider the ifelse() function from base R:

ifelse(3 > 5, 1, "this is false")
## [1] "this is false"

and compare it to {dplyr}’s implementation, if_else():

if_else(3 > 5, 1, "this is false")
Error: `false` must be type double, not character
Call `rlang::last_error()` to see a backtrace

if_else() fails because the return value when FALSE is not a double (a real number) but a character. This might seem unnecessarily strict, but at least it is predictable. This makes debugging easier when used inside functions. In Chapter 8 we are going to learn how to write our own functions, and being strict makes programming easier.

Basic string manipulation

For now, we have not closely studied character objects, we only learned how to define them. Later, in Chapter 5 we will learn about the {stringr} package which provides useful function to work with strings. However, there are several base R functions that are very useful that you might want to know nonetheless, such as paste() and paste0():

paste("Hello", "amigo")
## [1] "Hello amigo"

but you can also change the separator if needed:

paste("Hello", "amigo", sep = "--")
## [1] "Hello--amigo"

paste0() is the same as paste() but does not have any sep argument:

paste0("Hello", "amigo")
## [1] "Helloamigo"

If you provide a vector of characters, you can also use the collapse argument, which places whatever you provide for collapse between the characters of the vector:

paste0(c("Joseph", "Mary", "Jesus"), collapse = ", and ")
## [1] "Joseph, and Mary, and Jesus"

To change the case of characters, you can use toupper() and tolower():

## [1] "hahahahah"

Mathematical functions

Finally, there are the classical mathematical functions that you know and love:

  • sqrt()
  • exp()
  • log()
  • abs()
  • sin(), cos(), tan(), and others
  • sum(), cumsum(), prod(), cumprod()
  • max(), min()

and many others…

Hope you enjoyed! If you found this blog post useful, you might want to follow me on twitter for blog post updates and buy me an espresso or paypal.me.

Buy me an EspressoBuy me an Espresso