nest {tidyr}

Nest rows into a list-column of data frames

Description

Nesting creates a list-column of data frames; unnesting flattens it back out into regular columns. Nesting is implicitly a summarising operation: you get one row for each group defined by the non-nested columns. This is useful in conjunction with other summaries that work with whole datasets, most notably models.

Learn more in vignette("nest").

Usage

nest(.data, ..., .by = NULL, .key = NULL, .names_sep = NULL)

Arguments

.data

A data frame.

...

<tidy-select> Columns to nest; these will appear in the inner data frames.

Specified using name-variable pairs of the form new_col = c(col1, col2, col3). The right hand side can be any valid tidyselect expression.

If not supplied, then ... is derived as all columns not selected by .by, and will use the column name from .key.

: previously you could write df %>% nest(x, y, z). Convert to df %>% nest(data = c(x, y, z)).

.by

<tidy-select> Columns to nest by; these will remain in the outer data frame.

.by can be used in place of or in conjunction with columns supplied through ....

If not supplied, then .by is derived as all columns not selected by ....

.key

The name of the resulting nested column. Only applicable when ... isn't specified, i.e. in the case of df %>% nest(.by = x).

If NULL, then "data" will be used by default.

.names_sep

If NULL, the default, the inner names will come from the former outer names. If a string, the new inner names will use the outer names with names_sep automatically stripped. This makes names_sep roughly symmetric between nesting and unnesting.

Details

If neither ... nor .by are supplied, nest() will nest all variables, and will use the column name supplied through .key.

New syntax

tidyr 1.0.0 introduced a new syntax for nest() and unnest() that's designed to be more similar to other functions. Converting to the new syntax should be straightforward (guided by the message you'll receive) but if you just need to run an old analysis, you can easily revert to the previous behaviour using nest_legacy() and unnest_legacy() as follows:

library(tidyr)
nest <- nest_legacy
unnest <- unnest_legacy

Grouped data frames

df %>% nest(data = c(x, y)) specifies the columns to be nested; i.e. the columns that will appear in the inner data frame. df %>% nest(.by = c(x, y)) specifies the columns to nest by; i.e. the columns that will remain in the outer data frame. An alternative way to achieve the latter is to nest() a grouped data frame created by dplyr::group_by(). The grouping variables remain in the outer data frame and the others are nested. The result preserves the grouping of the input.

Variables supplied to nest() will override grouping variables so that df %>% group_by(x, y) %>% nest(data = !z) will be equivalent to df %>% nest(data = !z).

You can't supply .by with a grouped data frame, as the groups already represent what you are nesting by.

Examples

df <- tibble(x = c(1, 1, 1, 2, 2, 3), y = 1:6, z = 6:1)

# Specify variables to nest using name-variable pairs.
# Note that we get one row of output for each unique combination of
# non-nested variables.
df %>% nest(data = c(y, z))

# Specify variables to nest by (rather than variables to nest) using `.by`
df %>% nest(.by = x)

# In this case, since `...` isn't used you can specify the resulting column
# name with `.key`
df %>% nest(.by = x, .key = "cols")

# Use tidyselect syntax and helpers, just like in `dplyr::select()`
df %>% nest(data = any_of(c("y", "z")))

# `...` and `.by` can be used together to drop columns you no longer need,
# or to include the columns you are nesting by in the inner data frame too.
# This drops `z`:
df %>% nest(data = y, .by = x)
# This includes `x` in the inner data frame:
df %>% nest(data = everything(), .by = x)

# Multiple nesting structures can be specified at once
iris %>%
  nest(petal = starts_with("Petal"), sepal = starts_with("Sepal"))
iris %>%
  nest(width = contains("Width"), length = contains("Length"))

# Nesting a grouped data frame nests all variables apart from the group vars
fish_encounters %>%
  dplyr::group_by(fish) %>%
  nest()

# That is similar to `nest(.by = )`, except here the result isn't grouped
fish_encounters %>%
  nest(.by = fish)

# Nesting is often useful for creating per group models
mtcars %>%
  nest(.by = cyl) %>%
  dplyr::mutate(models = lapply(data, function(df) lm(mpg ~ wt, data = df)))