Source: R/join.R
mutate-joins.Rd
Mutating joins add columns from y
to x
, matching observations based onthe keys. There are four mutating joins: the inner join, and the three outerjoins.
Inner join
An inner_join()
only keeps observations from x
that have a matching keyin y
.
The most important property of an inner join is that unmatched rows in eitherinput are not included in the result. This means that generally inner joinsare not appropriate in most analyses, because it is too easy to loseobservations.
Outer joins
The three outer joins keep observations that appear in at least one of thedata frames:
A
left_join()
keeps all observations inx
.A
right_join()
keeps all observations iny
.A
full_join()
keeps all observations inx
andy
.
Usage
inner_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL)# S3 method for data.frameinner_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL, na_matches = c("na", "never"), multiple = "all", unmatched = "drop", relationship = NULL)left_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL)# S3 method for data.frameleft_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL, na_matches = c("na", "never"), multiple = "all", unmatched = "drop", relationship = NULL)right_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL)# S3 method for data.frameright_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL, na_matches = c("na", "never"), multiple = "all", unmatched = "drop", relationship = NULL)full_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL)# S3 method for data.framefull_join( x, y, by = NULL, copy = FALSE, suffix = c(".x", ".y"), ..., keep = NULL, na_matches = c("na", "never"), multiple = "all", relationship = NULL)
Arguments
- x, y
A pair of data frames, data frame extensions (e.g. a tibble), orlazy data frames (e.g. from dbplyr or dtplyr). See Methods, below, formore details.
- by
A join specification created with
join_by()
, or a charactervector of variables to join by.If
NULL
, the default,*_join()
will perform a natural join, using allvariables in common acrossx
andy
. A message lists the variables sothat you can check they're correct; suppress the message by supplyingby
explicitly.To join on different variables between
x
andy
, use ajoin_by()
specification. For example,join_by(a == b)
will matchx$a
toy$b
.To join by multiple variables, use a
join_by()
specification withmultiple expressions. For example,join_by(a == b, c == d)
will matchx$a
toy$b
andx$c
toy$d
. If the column names are the same betweenx
andy
, you can shorten this by listing only the variable names, likejoin_by(a, c)
.join_by()
can also be used to perform inequality, rolling, and overlapjoins. See the documentation at ?join_by for details onthese types of joins.See AlsoJoin in R: How to join (merge) data frames (inner, outer, left, right) in R - DataScience Made SimpleHow to merge data in R using R merge, dplyr, or data.tableR Merge - How To Merge Two R Data Frames - ProgrammingRMerge data frames in RFor simple equality joins, you can alternatively specify a character vectorof variable names to join by. For example,
by = c("a", "b")
joinsx$a
toy$a
andx$b
toy$b
. If variable names differ betweenx
andy
,use a named character vector likeby = c("x_a" = "y_a", "x_b" = "y_b")
.To perform a cross-join, generating all combinations of
x
andy
, seecross_join()
.- copy
If
x
andy
are not from the same data source,andcopy
isTRUE
, theny
will be copied into thesame src asx
. This allows you to join tables across srcs, butit is a potentially expensive operation so you must opt into it.- suffix
If there are non-joined duplicate variables in
x
andy
, these suffixes will be added to the output to disambiguate them.Should be a character vector of length 2.- ...
Other parameters passed onto methods.
- keep
Should the join keys from both
x
andy
be preserved in theoutput?If
NULL
, the default, joins on equality retain only the keys fromx
,while joins on inequality retain the keys from both inputs.If
TRUE
, all keys from both inputs are retained.If
FALSE
, only keys fromx
are retained. For right and full joins,the data in key columns corresponding to rows that only exist iny
aremerged into the key columns fromx
. Can't be used when joining oninequality conditions.
- na_matches
Should two
NA
or twoNaN
values match?- multiple
Handling of rows in
x
with multiple matches iny
.For each row ofx
:"all"
, the default, returns every match detected iny
. This is thesame behavior as SQL."any"
returns one match detected iny
, with no guarantees on whichmatch will be returned. It is often faster than"first"
and"last"
if you just need to detect if there is at least one match."first"
returns the first match detected iny
."last"
returns the last match detected iny
.
- unmatched
How should unmatched keys that would result in dropped rowsbe handled?
"drop"
drops unmatched keys from the result."error"
throws an error if unmatched keys are detected.
unmatched
is intended to protect you from accidentally dropping rowsduring a join. It only checks for unmatched keys in the input that couldpotentially drop rows.For left joins, it checks
y
.For right joins, it checks
x
.For inner joins, it checks both
x
andy
. In this case,unmatched
isalso allowed to be a character vector of length 2 to specify the behaviorforx
andy
independently.
- relationship
Handling of the expected relationship between the keys of
x
andy
. If the expectations chosen from the list below areinvalidated, an error is thrown.NULL
, the default, doesn't expect there to be any relationship betweenx
andy
. However, for equality joins it will check for a many-to-manyrelationship (which is typically unexpected) and will warn if one occurs,encouraging you to either take a closer look at your inputs or make thisrelationship explicit by specifying"many-to-many"
.See the Many-to-many relationships section for more details.
"one-to-one"
expects:Each row in
x
matches at most 1 row iny
.Each row in
y
matches at most 1 row inx
.
"one-to-many"
expects:Each row in
y
matches at most 1 row inx
.
"many-to-one"
expects:Each row in
x
matches at most 1 row iny
.
"many-to-many"
doesn't perform any relationship checks, but is providedto allow you to be explicit about this relationship if you know itexists.
relationship
doesn't handle cases where there are zero matches. For that,seeunmatched
.
Value
An object of the same type as x
(including the same groups). The order ofthe rows and columns of x
is preserved as much as possible. The output hasthe following properties:
The rows are affect by the join type.
inner_join()
returns matchedx
rows.left_join()
returns allx
rows.right_join()
returns matched ofx
rows, followed by unmatchedy
rows.full_join()
returns allx
rows, followed by unmatchedy
rows.
Output columns include all columns from
x
and all non-key columns fromy
. Ifkeep = TRUE
, the key columns fromy
are included as well.If non-key columns in
x
andy
have the same name,suffix
es are addedto disambiguate. Ifkeep = TRUE
and key columns inx
andy
havethe same name,suffix
es are added to disambiguate these as well.If
keep = FALSE
, output columns included inby
are coerced to theircommon type betweenx
andy
.
Many-to-many relationships
By default, dplyr guards against many-to-many relationships in equality joinsby throwing a warning. These occur when both of the following are true:
A row in
x
matches multiple rows iny
.A row in
y
matches multiple rows inx
.
This is typically surprising, as most joins involve a relationship ofone-to-one, one-to-many, or many-to-one, and is often the result of animproperly specified join. Many-to-many relationships are particularlyproblematic because they can result in a Cartesian explosion of the number ofrows returned from the join.
If a many-to-many relationship is expected, silence this warning byexplicitly setting relationship = "many-to-many"
.
In production code, it is best to preemptively set relationship
to whateverrelationship you expect to exist between the keys of x
and y
, as thisforces an error to occur immediately if the data doesn't align with yourexpectations.
Inequality joins typically result in many-to-many relationships by nature, sothey don't warn on them by default, but you should still take extra care whenspecifying an inequality join, because they also have the capability toreturn a large number of rows.
Rolling joins don't warn on many-to-many relationships either, but manyrolling joins follow a many-to-one relationship, so it is often useful toset relationship = "many-to-one"
to enforce this.
Note that in SQL, most database providers won't let you specify amany-to-many relationship between two tables, instead requiring that youcreate a third junction table that results in two one-to-many relationshipsinstead.
Methods
These functions are generics, which means that packages can provideimplementations (methods) for other classes. See the documentation ofindividual methods for extra arguments and differences in behaviour.
Methods available in currently loaded packages:
inner_join()
: dbplyr (tbl_lazy
), dplyr (data.frame
).left_join()
: dbplyr (tbl_lazy
), dplyr (data.frame
).right_join()
: dbplyr (tbl_lazy
), dplyr (data.frame
).full_join()
: dbplyr (tbl_lazy
), dplyr (data.frame
).
See also
Other joins: cross_join()
,filter-joins
,nest_join()
Examples
band_members %>% inner_join(band_instruments)#> Joining with `by = join_by(name)`#> # A tibble: 2 × 3#> name band plays #> <chr> <chr> <chr> #> 1 John Beatles guitar#> 2 Paul Beatles bass band_members %>% left_join(band_instruments)#> Joining with `by = join_by(name)`#> # A tibble: 3 × 3#> name band plays #> <chr> <chr> <chr> #> 1 Mick Stones NA #> 2 John Beatles guitar#> 3 Paul Beatles bass band_members %>% right_join(band_instruments)#> Joining with `by = join_by(name)`#> # A tibble: 3 × 3#> name band plays #> <chr> <chr> <chr> #> 1 John Beatles guitar#> 2 Paul Beatles bass #> 3 Keith NA guitarband_members %>% full_join(band_instruments)#> Joining with `by = join_by(name)`#> # A tibble: 4 × 3#> name band plays #> <chr> <chr> <chr> #> 1 Mick Stones NA #> 2 John Beatles guitar#> 3 Paul Beatles bass #> 4 Keith NA guitar# To suppress the message about joining variables, supply `by`band_members %>% inner_join(band_instruments, by = join_by(name))#> # A tibble: 2 × 3#> name band plays #> <chr> <chr> <chr> #> 1 John Beatles guitar#> 2 Paul Beatles bass # This is good practice in production code# Use an equality expression if the join variables have different namesband_members %>% full_join(band_instruments2, by = join_by(name == artist))#> # A tibble: 4 × 3#> name band plays #> <chr> <chr> <chr> #> 1 Mick Stones NA #> 2 John Beatles guitar#> 3 Paul Beatles bass #> 4 Keith NA guitar# By default, the join keys from `x` and `y` are coalesced in the output; use# `keep = TRUE` to keep the join keys from both `x` and `y`band_members %>% full_join(band_instruments2, by = join_by(name == artist), keep = TRUE)#> # A tibble: 4 × 4#> name band artist plays #> <chr> <chr> <chr> <chr> #> 1 Mick Stones NA NA #> 2 John Beatles John guitar#> 3 Paul Beatles Paul bass #> 4 NA NA Keith guitar# If a row in `x` matches multiple rows in `y`, all the rows in `y` will be# returned once for each matching row in `x`.df1 <- tibble(x = 1:3)df2 <- tibble(x = c(1, 1, 2), y = c("first", "second", "third"))df1 %>% left_join(df2)#> Joining with `by = join_by(x)`#> # A tibble: 4 × 2#> x y #> <dbl> <chr> #> 1 1 first #> 2 1 second#> 3 2 third #> 4 3 NA # If a row in `y` also matches multiple rows in `x`, this is known as a# many-to-many relationship, which is typically a result of an improperly# specified join or some kind of messy data. In this case, a warning is# thrown by default:df3 <- tibble(x = c(1, 1, 1, 3))df3 %>% left_join(df2)#> Joining with `by = join_by(x)`#> Warning: Detected an unexpected many-to-many relationship between `x` and `y`.#> ℹ Row 1 of `x` matches multiple rows in `y`.#> ℹ Row 1 of `y` matches multiple rows in `x`.#> ℹ If a many-to-many relationship is expected, set `relationship =#> "many-to-many"` to silence this warning.#> # A tibble: 7 × 2#> x y #> <dbl> <chr> #> 1 1 first #> 2 1 second#> 3 1 first #> 4 1 second#> 5 1 first #> 6 1 second#> 7 3 NA # In the rare case where a many-to-many relationship is expected, set# `relationship = "many-to-many"` to silence this warningdf3 %>% left_join(df2, relationship = "many-to-many")#> Joining with `by = join_by(x)`#> # A tibble: 7 × 2#> x y #> <dbl> <chr> #> 1 1 first #> 2 1 second#> 3 1 first #> 4 1 second#> 5 1 first #> 6 1 second#> 7 3 NA # Use `join_by()` with a condition other than `==` to perform an inequality# join. Here we match on every instance where `df1$x > df2$x`.df1 %>% left_join(df2, join_by(x > x))#> # A tibble: 6 × 3#> x.x x.y y #> <int> <dbl> <chr> #> 1 1 NA NA #> 2 2 1 first #> 3 2 1 second#> 4 3 1 first #> 5 3 1 second#> 6 3 2 third # By default, NAs match other NAs so that there are two# rows in the output of this join:df1 <- data.frame(x = c(1, NA), y = 2)df2 <- data.frame(x = c(1, NA), z = 3)left_join(df1, df2)#> Joining with `by = join_by(x)`#> x y z#> 1 1 2 3#> 2 NA 2 3# You can optionally request that NAs don't match, giving a# a result that more closely resembles SQL joinsleft_join(df1, df2, na_matches = "never")#> Joining with `by = join_by(x)`#> x y z#> 1 1 2 3#> 2 NA 2 NA