Grouped data

dplyr verbs are particularly powerful when you apply them to grouped data frames (grouped_df objects). This vignette shows you:

We’ll start by loading dplyr:

library(dplyr)

group_by()

The most important grouping verb is group_by(): it takes a data frame and one or more variables to group by:

by_species <- starwars %>% group_by(species)
by_sex_gender <- starwars %>% group_by(sex, gender)

You can see the grouping when you print the data:

by_species
#> # A tibble: 87 x 14
#> # Groups:   species [38]
#>   name  height  mass hair_color skin_color eye_color birth_year sex   gender
#>   <chr>  <int> <dbl> <chr>      <chr>      <chr>          <dbl> <chr> <chr> 
#> 1 Luke…    172    77 blond      fair       blue            19   male  mascu…
#> 2 C-3PO    167    75 <NA>       gold       yellow         112   none  mascu…
#> 3 R2-D2     96    32 <NA>       white, bl… red             33   none  mascu…
#> 4 Dart…    202   136 none       white      yellow          41.9 male  mascu…
#> # … with 83 more rows, and 5 more variables: homeworld <chr>, species <chr>,
#> #   films <list>, vehicles <list>, starships <list>
by_sex_gender
#> # A tibble: 87 x 14
#> # Groups:   sex, gender [6]
#>   name  height  mass hair_color skin_color eye_color birth_year sex   gender
#>   <chr>  <int> <dbl> <chr>      <chr>      <chr>          <dbl> <chr> <chr> 
#> 1 Luke…    172    77 blond      fair       blue            19   male  mascu…
#> 2 C-3PO    167    75 <NA>       gold       yellow         112   none  mascu…
#> 3 R2-D2     96    32 <NA>       white, bl… red             33   none  mascu…
#> 4 Dart…    202   136 none       white      yellow          41.9 male  mascu…
#> # … with 83 more rows, and 5 more variables: homeworld <chr>, species <chr>,
#> #   films <list>, vehicles <list>, starships <list>

Or use tally() to count the number of rows in each group. The sort argument is useful if you want to see the largest groups up front.

by_species %>% tally()
#> # A tibble: 38 x 2
#>   species      n
#>   <chr>    <int>
#> 1 Aleena       1
#> 2 Besalisk     1
#> 3 Cerean       1
#> 4 Chagrian     1
#> # … with 34 more rows

by_sex_gender %>% tally(sort = TRUE)
#> # A tibble: 6 x 3
#> # Groups:   sex [5]
#>   sex    gender        n
#>   <chr>  <chr>     <int>
#> 1 male   masculine    60
#> 2 female feminine     16
#> 3 none   masculine     5
#> 4 <NA>   <NA>          4
#> # … with 2 more rows

As well as grouping by existing variables, you can group by any function of existing variables. This is equivalent to performing a mutate() before the group_by():

bmi_breaks <- c(0, 18.5, 25, 30, Inf)

starwars %>%
  group_by(bmi_cat = cut(mass/(height/100)^2, breaks=bmi_breaks)) %>%
  tally()
#> # A tibble: 5 x 2
#>   bmi_cat       n
#>   <fct>     <int>
#> 1 (0,18.5]     10
#> 2 (18.5,25]    24
#> 3 (25,30]      13
#> 4 (30,Inf]     12
#> # … with 1 more row

Group metadata

You can see underlying group data with group_keys(). It has one row for each group and one column for each grouping variable:

by_species %>% group_keys()
#> # A tibble: 38 x 1
#>   species 
#>   <chr>   
#> 1 Aleena  
#> 2 Besalisk
#> 3 Cerean  
#> 4 Chagrian
#> # … with 34 more rows

by_sex_gender %>% group_keys()
#> # A tibble: 6 x 2
#>   sex            gender   
#>   <chr>          <chr>    
#> 1 female         feminine 
#> 2 hermaphroditic masculine
#> 3 male           masculine
#> 4 none           feminine 
#> # … with 2 more rows

You can see which group each row belongs to with group_indices():

by_species %>% group_indices()
#>  [1] 11  6  6 11 11 11 11  6 11 11 11 11 34 11 24 12 11 11 36 11 11  6 31 11 11
#> [26] 18 11 11  8 26 11 21 11 10 10 10 38 30  7 38 11 37 32 32 33 35 29 11  3 20
#> [51] 37 27 13 23 16  4 11 11 11  9 17 17 11 11 11 11  5  2 15 15 11  1  6 25 19
#> [76] 28 14 34 11 38 22 11 11 11  6 38 11

And which rows each group contains with group_rows():

by_species %>% group_rows() %>% head()
#> <list_of<integer>[6]>
#> [[1]]
#> [1] 72
#> 
#> [[2]]
#> [1] 68
#> 
#> [[3]]
#> [1] 49
#> 
#> [[4]]
#> [1] 56
#> 
#> [[5]]
#> [1] 67
#> 
#> [[6]]
#> [1]  2  3  8 22 73 85

Use group_vars() if you just want the names of the grouping variables:

by_species %>% group_vars()
#> [1] "species"
by_sex_gender %>% group_vars()
#> [1] "sex"    "gender"

Changing and adding to grouping variables

If you apply group_by() to an already grouped dataset, will overwrite the existing grouping variables. For example, the following code groups by homeworld instead of species:

by_species %>%
  group_by(homeworld) %>%
  tally()
#> # A tibble: 49 x 2
#>   homeworld       n
#>   <chr>       <int>
#> 1 Alderaan        3
#> 2 Aleen Minor     1
#> 3 Bespin          1
#> 4 Bestine IV      1
#> # … with 45 more rows

To augment the grouping, using .add = TRUE1. For example, the following code groups by species and homeworld:

by_species %>%
  group_by(homeworld, .add = TRUE) %>%
  tally()
#> # A tibble: 58 x 3
#> # Groups:   species [38]
#>   species  homeworld       n
#>   <chr>    <chr>       <int>
#> 1 Aleena   Aleen Minor     1
#> 2 Besalisk Ojom            1
#> 3 Cerean   Cerea           1
#> 4 Chagrian Champala        1
#> # … with 54 more rows

Removing grouping variables

To remove all grouping variables, use ungroup():

by_species %>%
  ungroup() %>%
  tally()
#> # A tibble: 1 x 1
#>       n
#>   <int>
#> 1    87

You can also choose to selectively ungroup by listing the variables you want to remove:

by_sex_gender %>% 
  ungroup(sex) %>% 
  tally()
#> # A tibble: 3 x 2
#>   gender        n
#>   <chr>     <int>
#> 1 feminine     17
#> 2 masculine    66
#> 3 <NA>          4

Verbs

The following sections describe how grouping affects the main dplyr verbs.

summarise()

summarise() computes a summary for each group. This means that it starts from group_keys(), adding summary variables to the right hand side:

by_species %>%
  summarise(
    n = n(),
    height = mean(height, na.rm = TRUE)
  )
#> # A tibble: 38 x 3
#>   species      n height
#>   <chr>    <int>  <dbl>
#> 1 Aleena       1     79
#> 2 Besalisk     1    198
#> 3 Cerean       1    198
#> 4 Chagrian     1    196
#> # … with 34 more rows

The .groups= argument controls the grouping structure of the output. The historical behaviour of removing the right hand side grouping variable corresponds to .groups = "drop_last" without a message or .groups = NULL with a message (the default).

by_sex_gender %>% 
  summarise(n = n()) %>% 
  group_vars()
#> `summarise()` has grouped output by 'sex'. You can override using the `.groups` argument.
#> [1] "sex"

by_sex_gender %>% 
  summarise(n = n(), .groups = "drop_last") %>% 
  group_vars()
#> [1] "sex"

Since version 1.0.0 the groups may also be kept (.groups = "keep") or dropped (.groups = "drop").

by_sex_gender %>% 
  summarise(n = n(), .groups = "keep") %>% 
  group_vars()
#> [1] "sex"    "gender"

by_sex_gender %>% 
  summarise(n = n(), .groups = "drop") %>% 
  group_vars()
#> character(0)

When the output no longer have grouping variables, it becomes ungrouped (i.e. a regular tibble).

select(), rename(), and relocate()

rename() and relocate() behave identically with grouped and ungrouped data because they only affect the name or position of existing columns. Grouped select() is almost identical to ungrouped select, except that it always includes the grouping variables:

by_species %>% select(mass)
#> Adding missing grouping variables: `species`
#> # A tibble: 87 x 2
#> # Groups:   species [38]
#>   species  mass
#>   <chr>   <dbl>
#> 1 Human      77
#> 2 Droid      75
#> 3 Droid      32
#> 4 Human     136
#> # … with 83 more rows

If you don’t want the grouping variables, you’ll have to first ungroup(). (This design is possibly a mistake, but we’re stuck with it for now.)

arrange()

Grouped arrange() is the same as ungrouped arrange(), unless you set .by_group = TRUE, in which case it will order first by the grouping variables.

by_species %>%
  arrange(desc(mass)) %>%
  relocate(species, mass)
#> # A tibble: 87 x 14
#> # Groups:   species [38]
#>   species  mass name  height hair_color skin_color eye_color birth_year sex  
#>   <chr>   <dbl> <chr>  <int> <chr>      <chr>      <chr>          <dbl> <chr>
#> 1 Hutt     1358 Jabb…    175 <NA>       green-tan… orange         600   herm…
#> 2 Kaleesh   159 Grie…    216 none       brown, wh… green, y…       NA   male 
#> 3 Droid     140 IG-88    200 none       metal      red             15   none 
#> 4 Human     136 Dart…    202 none       white      yellow          41.9 male 
#> # … with 83 more rows, and 5 more variables: gender <chr>, homeworld <chr>,
#> #   films <list>, vehicles <list>, starships <list>

by_species %>%
  arrange(desc(mass), .by_group = TRUE) %>%
  relocate(species, mass)
#> # A tibble: 87 x 14
#> # Groups:   species [38]
#>   species  mass name  height hair_color skin_color eye_color birth_year sex  
#>   <chr>   <dbl> <chr>  <int> <chr>      <chr>      <chr>          <dbl> <chr>
#> 1 Aleena     15 Ratt…     79 none       grey, blue unknown           NA male 
#> 2 Besali…   102 Dext…    198 none       brown      yellow            NA male 
#> 3 Cerean     82 Ki-A…    198 white      pale       yellow            92 male 
#> 4 Chagri…    NA Mas …    196 none       blue       blue              NA male 
#> # … with 83 more rows, and 5 more variables: gender <chr>, homeworld <chr>,
#> #   films <list>, vehicles <list>, starships <list>

Note that second example is sorted by species (from the group_by() statement) and then by mass (within species).

mutate() and transmute()

In simple cases with vectorised functions, grouped and ungrouped mutate() give the same results. They differ when used with summary functions:

# Subtract off global mean
starwars %>% 
  select(name, homeworld, mass) %>% 
  mutate(standard_mass = mass - mean(mass, na.rm = TRUE))
#> # A tibble: 87 x 4
#>   name           homeworld  mass standard_mass
#>   <chr>          <chr>     <dbl>         <dbl>
#> 1 Luke Skywalker Tatooine     77         -20.3
#> 2 C-3PO          Tatooine     75         -22.3
#> 3 R2-D2          Naboo        32         -65.3
#> 4 Darth Vader    Tatooine    136          38.7
#> # … with 83 more rows

# Subtract off homeworld mean
starwars %>% 
  select(name, homeworld, mass) %>% 
  group_by(homeworld) %>% 
  mutate(standard_mass = mass - mean(mass, na.rm = TRUE))
#> # A tibble: 87 x 4
#> # Groups:   homeworld [49]
#>   name           homeworld  mass standard_mass
#>   <chr>          <chr>     <dbl>         <dbl>
#> 1 Luke Skywalker Tatooine     77         -8.38
#> 2 C-3PO          Tatooine     75        -10.4 
#> 3 R2-D2          Naboo        32        -32.2 
#> 4 Darth Vader    Tatooine    136         50.6 
#> # … with 83 more rows

Or with window functions like min_rank():

# Overall rank
starwars %>% 
  select(name, homeworld, height) %>% 
  mutate(rank = min_rank(height))
#> # A tibble: 87 x 4
#>   name           homeworld height  rank
#>   <chr>          <chr>      <int> <int>
#> 1 Luke Skywalker Tatooine     172    29
#> 2 C-3PO          Tatooine     167    21
#> 3 R2-D2          Naboo         96     5
#> 4 Darth Vader    Tatooine     202    72
#> # … with 83 more rows

# Rank per homeworld
starwars %>% 
  select(name, homeworld, height) %>% 
  group_by(homeworld) %>% 
  mutate(rank = min_rank(height))
#> # A tibble: 87 x 4
#> # Groups:   homeworld [49]
#>   name           homeworld height  rank
#>   <chr>          <chr>      <int> <int>
#> 1 Luke Skywalker Tatooine     172     5
#> 2 C-3PO          Tatooine     167     4
#> 3 R2-D2          Naboo         96     1
#> 4 Darth Vader    Tatooine     202    10
#> # … with 83 more rows

filter()

A grouped filter() effectively does a mutate() to generate a logical variable, and then only keeps the rows where the variable is TRUE. This means that grouped filters can be used with summary functions. For example, we can find the tallest character of each species:

by_species %>%
  select(name, species, height) %>% 
  filter(height == max(height))
#> # A tibble: 35 x 3
#> # Groups:   species [35]
#>   name                  species        height
#>   <chr>                 <chr>           <int>
#> 1 Greedo                Rodian            173
#> 2 Jabba Desilijic Tiure Hutt              175
#> 3 Yoda                  Yoda's species     66
#> 4 Bossk                 Trandoshan        190
#> # … with 31 more rows

You can also use filter() to remove entire groups. For example, the following code eliminates all groups that only have a single member:

by_species %>%
  filter(n() != 1) %>% 
  tally()
#> # A tibble: 9 x 2
#>   species      n
#>   <chr>    <int>
#> 1 Droid        6
#> 2 Gungan       3
#> 3 Human       35
#> 4 Kaminoan     2
#> # … with 5 more rows

slice() and friends

slice() and friends (slice_head(), slice_tail(), slice_sample(), slice_min() and slice_max()) select rows within a group. For example, we can select the first observation within each species:

by_species %>%
  relocate(species) %>% 
  slice(1)
#> # A tibble: 38 x 14
#> # Groups:   species [38]
#>   species name  height  mass hair_color skin_color eye_color birth_year sex  
#>   <chr>   <chr>  <int> <dbl> <chr>      <chr>      <chr>          <dbl> <chr>
#> 1 Aleena  Ratt…     79    15 none       grey, blue unknown           NA male 
#> 2 Besali… Dext…    198   102 none       brown      yellow            NA male 
#> 3 Cerean  Ki-A…    198    82 white      pale       yellow            92 male 
#> 4 Chagri… Mas …    196    NA none       blue       blue              NA male 
#> # … with 34 more rows, and 5 more variables: gender <chr>, homeworld <chr>,
#> #   films <list>, vehicles <list>, starships <list>

Similarly, we can use slice_min() to select the smallest n values of a variable:

by_species %>%
  filter(!is.na(height)) %>% 
  slice_min(height, n = 2)
#> # A tibble: 48 x 14
#> # Groups:   species [38]
#>   name  height  mass hair_color skin_color eye_color birth_year sex   gender
#>   <chr>  <int> <dbl> <chr>      <chr>      <chr>          <dbl> <chr> <chr> 
#> 1 Ratt…     79    15 none       grey, blue unknown           NA male  mascu…
#> 2 Dext…    198   102 none       brown      yellow            NA male  mascu…
#> 3 Ki-A…    198    82 white      pale       yellow            92 male  mascu…
#> 4 Mas …    196    NA none       blue       blue              NA male  mascu…
#> # … with 44 more rows, and 5 more variables: homeworld <chr>, species <chr>,
#> #   films <list>, vehicles <list>, starships <list>

Computing on grouping information

Inside dplyr verbs, you can access various properties of the “current” group using a family of functions with the cur_ prefix. These functions are typically needed for everyday usage of dplyr, but can be useful because they allow you to free from some of the typical constraints of dplyr verbs.

cur_data()

cur_data() returns the current group, excluding grouping variables. It’s useful to feed to functions that take a whole data frame. For example, the following code fits a linear model of mass ~ height to each species:

by_species %>%
  filter(n() > 1) %>% 
  mutate(mod = list(lm(mass ~ height, data = cur_data())))
#> # A tibble: 58 x 15
#> # Groups:   species [9]
#>   name  height  mass hair_color skin_color eye_color birth_year sex   gender
#>   <chr>  <int> <dbl> <chr>      <chr>      <chr>          <dbl> <chr> <chr> 
#> 1 Luke…    172    77 blond      fair       blue            19   male  mascu…
#> 2 C-3PO    167    75 <NA>       gold       yellow         112   none  mascu…
#> 3 R2-D2     96    32 <NA>       white, bl… red             33   none  mascu…
#> 4 Dart…    202   136 none       white      yellow          41.9 male  mascu…
#> # … with 54 more rows, and 6 more variables: homeworld <chr>, species <chr>,
#> #   films <list>, vehicles <list>, starships <list>, mod <list>

cur_group() and cur_group_id()

cur_group_id() gives a unique numeric identifier for the current group. This is sometimes useful if you want to index into an external data structure.

by_species %>%
  arrange(species) %>% 
  select(name, species, homeworld) %>% 
  mutate(id = cur_group_id())
#> # A tibble: 87 x 4
#> # Groups:   species [38]
#>   name            species  homeworld      id
#>   <chr>           <chr>    <chr>       <int>
#> 1 Ratts Tyerell   Aleena   Aleen Minor     1
#> 2 Dexter Jettster Besalisk Ojom            2
#> 3 Ki-Adi-Mundi    Cerean   Cerea           3
#> 4 Mas Amedda      Chagrian Champala        4
#> # … with 83 more rows

  1. Note that the argument changed from add = TRUE to .add = TRUE in dplyr 1.0.0.↩︎