Introduction to R

• We'll start by using the console as a calculator

• Results of commands are returned there

2 + 3

[1] 5

• Results can be stored as objects (AKA variables)

a <- 2 + 3
a

[1] 5

• Character strings can also be stored as data

string <- "Hello world"
string

[1] "Hello world"

• Vectors hold multiple elements of the same type, can be used for element-wise operations

x <- c(2, 18, 12, 23, 73)
x - 3

[1] -1 15  9 20 70

str(x)

 num [1:5] 2 18 12 23 73

• The c function “c"oerces elements into a vector
• The str function lets us see the "str"ucture of the data
• Vectors can also contain character strings and logical values

c("Learning", "R", "is", "fun!")

[1] "Learning" "R"        "is"       "fun!"    

c(T, F, F, F, T, F)

[1]  TRUE FALSE FALSE FALSE  TRUE FALSE

• Values in vectors (and other data types) can be missing
• NA is the most important missing value
  • Used for both numeric and text data

b <- c(1, 2, 3, NA, 5)
is.na(b)

[1] FALSE FALSE FALSE  TRUE FALSE

• Math functions also return Inf, -Inf, and NaN

• You can reference cells of vectors using [n] function/notation
  • n is a number or vector of numbers/logical values

x <- c("Learning", "R", "is", "fun!")
x[1]

[1] "Learning"

x[c(1, 4)]

[1] "Learning" "fun!"    

x[c(T, F, F, T)]

[1] "Learning" "fun!"    

• Element-wise operations on vectors work for all kinds of data

v <- c("abcd", "efgh", "ijkl")
v

[1] "abcd" "efgh" "ijkl"

substr(v, 1, 2)

[1] "ab" "ef" "ij"

• However, R can't give meaning to character strings on its own

f <- c("Red", "Blue", "Red", "Blue")
summary(f)

   Length     Class      Mode 
        4 character character 

• We can create a factor, which creates values according to category

f <- factor(f)
summary(f)

Blue  Red 
   2    2 

• Factors are similar to using labels for categories in Stata
• R can also order the levels in factors

temperatures <- c("High", "Low", "High","Low", "Medium")
temp_factor <- factor(temperatures, 
                      order = TRUE, 
                      levels = c("Low", "Medium", "High"))
temp_factor

[1] High   Low    High   Low    Medium
Levels: Low < Medium < High

• Matrices are like two-dimensional vectors
  • Can only hold one type of data

matrix(1:9, byrow = TRUE, nrow = 3)

     [,1] [,2] [,3]
[1,]    1    2    3
[2,]    4    5    6
[3,]    7    8    9

• Note that 1:9 generates a vector of all integers from 1 through 9

• You can create matrices from vectors

col1 <- c(1, 2, 3)
col2 <- c(4, 5, 6)
col3 <- c(7, 8, 9)
matrix(c(col1, col2, col3), ncol = 3)

     [,1] [,2] [,3]
[1,]    1    4    7
[2,]    2    5    8
[3,]    3    6    9

• Refer to cells using [ ] and two dimensions
  • First term for rows, second term for columns

m <- matrix(c(col1, col2, col3), ncol = 3)
m

     [,1] [,2] [,3]
[1,]    1    4    7
[2,]    2    5    8
[3,]    3    6    9

m[2, 2]

[1] 5

• Similar to a vector, but can contain multiple data types in on object, even other lists
• Not critical for beginners, but new object types from packages are often built on lists

l <- list(a, v, matrix(1:9, byrow = TRUE, nrow = 3))
str(l)

List of 3
 $ : num 5
 $ : chr [1:3] "abcd" "efgh" "ijkl"
 $ : int [1:3, 1:3] 1 4 7 2 5 8 3 6 9

• Reference list elements with double brackets [[ ]]

l[[2]]

[1] "abcd" "efgh" "ijkl"

• Add [ ] to reference cells of object in list

l[[3]][2, 2]

[1] 5

• In my view, matrices are best used for math calculations
• Restrictive in usage, hard to create and manipulate
• You will most likely be using data frames in R

data.frame(col1, col2, col3)

  col1 col2 col3
1    1    4    7
2    2    5    8
3    3    6    9

• Data frames store a variety of data types in columns of the same table
  • Like a standard dataset in Stata
  • However, you can load as many data frames as you want!!!

col2 <- c("lmao", "brb", "smh")            #character
col3 <- factor(c("Good", "Good", "Bad"))   #factor
col4 <- c(T, F, T)                         #logical
df <- data.frame(col1, col2, col3, col4, stringsAsFactors = F)

df

  col1 col2 col3  col4
1    1 lmao Good  TRUE
2    2  brb Good FALSE
3    3  smh  Bad  TRUE

summary(df)

      col1         col2             col3      col4        
 Min.   :1.0   Length:3           Bad :1   Mode :logical  
 1st Qu.:1.5   Class :character   Good:2   FALSE:1        
 Median :2.0   Mode  :character            TRUE :2        
 Mean   :2.0                                              
 3rd Qu.:2.5                                              
 Max.   :3.0                                              

• You can reference data frames like matrices with [ ]
• Reference and add columns by name with $

df$new <- 0
df$new

[1] 0 0 0

df

  col1 col2 col3  col4 new
1    1 lmao Good  TRUE   0
2    2  brb Good FALSE   0
3    3  smh  Bad  TRUE   0

• You can load data from a range of file types with built-in functions
• We'll look at packages that cover even more

#change to own file path
library(here)
salaries <- read.csv(here("data","white-house-salaries.csv"))
summary(salaries)

             employee_name      salary      
 brooke, mary j     :  16   Min.   :     0  
 campbell, frances l:  16   1st Qu.: 45000  
 droege, philip c   :  16   Median : 61952  
 jones, crystal b   :  16   Mean   : 76412  
 kalbaugh, david e  :  16   3rd Qu.:102000  
 mattson, philip c  :  16   Max.   :225000  
 (Other)            :7012                   
                       position         year     
 staff assistant           : 429   Min.   :2001  
 associate director        : 221   1st Qu.:2006  
 records management analyst: 206   Median :2010  
 deputy associate director : 176   Mean   :2010  
 executive assistant       : 154   3rd Qu.:2014  
 (Other)                   :5920   Max.   :2017  
 NA's                      :   2                 
                  status            party      president        term     
 detailee            : 328   democrat  :3740   bush :2991   first :3982  
 employee            :6778   republican:3368   obama:3740   second:3126  
 employee (part-time):   2                     trump: 377                




    gender    
 female:3521  
 male  :3587