Two-table verbs

It’s rare that a data analysis involves only a single table of data. In practice, you’ll normally have many tables that contribute to an analysis, and you need flexible tools to combine them. In dplyr, there are three families of verbs that work with two tables at a time:

(This discussion assumes that you have tidy data, where the rows are observations and the columns are variables. If you’re not familiar with that framework, I’d recommend reading up on it first.)

All two-table verbs work similarly. The first two arguments are x and y, and provide the tables to combine. The output is always a new table with the same type as x.

Mutating joins

Mutating joins allow you to combine variables from multiple tables. For example, take the nycflights13 data. In one table we have flight information with an abbreviation for carrier, and in another we have a mapping between abbreviations and full names. You can use a join to add the carrier names to the flight data:

library("nycflights13")
# Drop unimportant variables so it's easier to understand the join results.
flights2 <- flights %>% select(year:day, hour, origin, dest, tailnum, carrier)

flights2 %>% 
  left_join(airlines)
#> Joining, by = "carrier"
#> # A tibble: 336,776 x 9
#>    year month   day  hour origin dest  tailnum carrier name                  
#>   <int> <int> <int> <dbl> <chr>  <chr> <chr>   <chr>   <chr>                 
#> 1  2013     1     1     5 EWR    IAH   N14228  UA      United Air Lines Inc. 
#> 2  2013     1     1     5 LGA    IAH   N24211  UA      United Air Lines Inc. 
#> 3  2013     1     1     5 JFK    MIA   N619AA  AA      American Airlines Inc.
#> 4  2013     1     1     5 JFK    BQN   N804JB  B6      JetBlue Airways       
#> 5  2013     1     1     6 LGA    ATL   N668DN  DL      Delta Air Lines Inc.  
#> # … with 336,771 more rows

Controlling how the tables are matched

As well as x and y, each mutating join takes an argument by that controls which variables are used to match observations in the two tables. There are a few ways to specify it, as I illustrate below with various tables from nycflights13:

Types of join

There are four types of mutating join, which differ in their behaviour when a match is not found. We’ll illustrate each with a simple example:

df1 <- tibble(x = c(1, 2), y = 2:1)
df2 <- tibble(x = c(3, 1), a = 10, b = "a")

The left, right and full joins are collectively know as outer joins. When a row doesn’t match in an outer join, the new variables are filled in with missing values.

Observations

While mutating joins are primarily used to add new variables, they can also generate new observations. If a match is not unique, a join will add all possible combinations (the Cartesian product) of the matching observations:

df1 <- tibble(x = c(1, 1, 2), y = 1:3)
df2 <- tibble(x = c(1, 1, 2), z = c("a", "b", "a"))

df1 %>% left_join(df2)
#> Joining, by = "x"
#> # A tibble: 5 x 3
#>       x     y z    
#>   <dbl> <int> <chr>
#> 1     1     1 a    
#> 2     1     1 b    
#> 3     1     2 a    
#> 4     1     2 b    
#> 5     2     3 a

Filtering joins

Filtering joins match observations in the same way as mutating joins, but affect the observations, not the variables. There are two types:

These are most useful for diagnosing join mismatches. For example, there are many flights in the nycflights13 dataset that don’t have a matching tail number in the planes table:

library("nycflights13")
flights %>% 
  anti_join(planes, by = "tailnum") %>% 
  count(tailnum, sort = TRUE)
#> # A tibble: 722 x 2
#>   tailnum     n
#>   <chr>   <int>
#> 1 <NA>     2512
#> 2 N725MQ    575
#> 3 N722MQ    513
#> 4 N723MQ    507
#> 5 N713MQ    483
#> # … with 717 more rows

If you’re worried about what observations your joins will match, start with a semi_join() or anti_join(). semi_join() and anti_join() never duplicate; they only ever remove observations.

df1 <- tibble(x = c(1, 1, 3, 4), y = 1:4)
df2 <- tibble(x = c(1, 1, 2), z = c("a", "b", "a"))

# Four rows to start with:
df1 %>% nrow()
#> [1] 4
# And we get four rows after the join
df1 %>% inner_join(df2, by = "x") %>% nrow()
#> [1] 4
# But only two rows actually match
df1 %>% semi_join(df2, by = "x") %>% nrow()
#> [1] 2

Set operations

The final type of two-table verb is set operations. These expect the x and y inputs to have the same variables, and treat the observations like sets:

Given this simple data:

(df1 <- tibble(x = 1:2, y = c(1L, 1L)))
#> # A tibble: 2 x 2
#>       x     y
#>   <int> <int>
#> 1     1     1
#> 2     2     1
(df2 <- tibble(x = 1:2, y = 1:2))
#> # A tibble: 2 x 2
#>       x     y
#>   <int> <int>
#> 1     1     1
#> 2     2     2

The four possibilities are:

intersect(df1, df2)
#> # A tibble: 1 x 2
#>       x     y
#>   <int> <int>
#> 1     1     1
# Note that we get 3 rows, not 4
union(df1, df2)
#> # A tibble: 3 x 2
#>       x     y
#>   <int> <int>
#> 1     1     1
#> 2     2     1
#> 3     2     2
setdiff(df1, df2)
#> # A tibble: 1 x 2
#>       x     y
#>   <int> <int>
#> 1     2     1
setdiff(df2, df1)
#> # A tibble: 1 x 2
#>       x     y
#>   <int> <int>
#> 1     2     2

Multiple-table verbs

dplyr does not provide any functions for working with three or more tables. Instead use purrr::reduce() or Reduce(), as described in Advanced R, to iteratively combine the two-table verbs to handle as many tables as you need.