LECTURE 4: DATA STRUCTURES IN R (contd)

STAT598z: Intro. to computing for statistics

---

Vinayak Rao

Department of Statistics, Purdue University

Data frames

Very common and convenient data structures

Used to store tables:

• Columns are variables and rows are observations

​	Age	PhD	GPA
Alice	25	TRUE	3.6
Bob	24	TRUE	3.4
Carol	21	FALSE	3.8

An R data frame is a list of equal length vectors

df <- data.frame(age = c(25L,24L,21L),  # Warning: df is an
                 PhD = c( T , T , F ),  #   R function
                 GPA = c(3.6,2.4,2.8))

print(df)

typeof(df)

class(df)

Since data frames are lists, we can use list indexing

Can also use matrix indexing (more convenient)

print(df[2,'age'])

print(df[2,])

print(df$GPA)

nrow(df)*ncol(df)

list functions apply as usual

matrix functions are also interpreted intuitively

Useful functions are:

• 'length(), dim(), nrow(), ncol()'
• 'names()' (or 'colnames()')', rownames'
• 'rbind(), cbind()'

rownames(df) <- c("Alice", "Bob", "Carol")

df[4,1] <- 30L; print(df)

Many R datasets are data frames

library("datasets")
class(mtcars)

print(head(mtcars)) # Print part of a large object

Tibbles

Tibbles are essentially dataframes, with some convenience features

Interact will with the tidyverse package (later)

library(tidyverse)
t_mt_cars <- as_tibble(mtcars)
class(t_mt_cars)

Tribbles print more nicely that dataframes (but see RStudio's View())

print(t_mt_cars)

You can reference columns of a tibble as you create it

sin_tb <- tibble(x=seq(-5,5,.1), y=sin(x));
print(sin_tb)

Factors

Categorical variables that take on a finite number of values

• Employee type: student/staff/faculty
• Grade: A/B/C/F

Useful when variable can take a fixed set of values (unlike character strings)

R implements these internally as integer vectors

Has two attributes to distinguish from regular integers:

levels() specifies possible values the factor can take

• E.g. c("male", "female")

class = factor tells R to check for violations

# Character vector for 4 students
grades_bad <- c("a", "a", "b", "f")

# Factor vector for 4 students
grades <- factor(c("a", "a", "b", "f"))

print(grades);

typeof(grades)

class(grades)

levels(grades) # Not quite what we wanted!

grades <- factor(c("a", "a", "b", "f"))
str(grades)

grades[2] <- "c"

str(grades)

grades <- factor(c("a","a","b","a","f"),
             levels = c("a","b","c","f"))

str(grades)

table(grades)   # table also works with other data-types

Factors can be ordered:

grades <- factor(c("a","a","b","f"),
            levels = c("f","c","b","a"),
            ordered = TRUE )
grades

grades[1] > grades[3]

gl(): Generate factors levels

Usage (from the R documentation):

gl(n, k, length = n * k, labels = seq_len(n),
   ordered = FALSE )

Look at the examples there:

# First control, then treatment:
gl(2, 8, labels = c("Control", "Treat"))

gl(2, 1, 20) # 20 alternating 1s and 2s

gl(2, 2, 20) # alternating pairs of 1s and 2s