In hadley/tibble: Simple Data Frames
#| include: false
# To suppress messages
library(tibble)
library(vctrs)
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>",
error = TRUE
)
tibble:::set_dftbl_hooks()
options(
lifecycle_verbosity = "warning",
lifecycle_disable_warnings = FALSE,
lifecycle_verbose_soft_deprecation = TRUE,
lifecycle_repeat_warnings = TRUE
)
# Set to FALSE for production
eval_details <- (Sys.getenv("IN_GALLEY") != "")
This vignette defines invariants for subsetting and subset-assignment for tibbles, and illustrates where their behaviour differs from data frames.
The goal is to define a small set of invariants that consistently define how behaviors interact.
Some behaviors are defined using functions of the vctrs package, e.g. vec_slice(), vec_recycle() and vec_as_index().
Refer to their documentation for more details about the invariants that they follow.
The subsetting and subassignment operators for data frames and tibbles are particularly tricky, because they support both row and column indexes, both of which are optionally missing.
We resolve this by first defining column access with [[ and $, then column-wise subsetting with [, then row-wise subsetting, then the composition of both.
Conventions
In this article, all behaviors are demonstrated using one example data frame and its tibble equivalent:
#| label: setup
library(tibble)
library(vctrs)
new_df <- function() {
df <- data.frame(n = c(1L, NA, 3L, NA))
df$c <- letters[5:8]
df$li <- list(9, 10:11, 12:14, "text")
df
}
new_tbl <- function() {
as_tibble(new_df())
}
Results of the same code for data frames and tibbles are presented side by side:
#| label: show
#| dftbl: true
#| dftbl_always: true
new_df()
If the results are identical (after converting to a data frame if necessary), only the tibble result is shown.
Subsetting operations are read-only.
The same objects are reused in all examples:
#| label: ro
df <- new_df()
tbl <- new_tbl()
Where needed, we also show examples with hierarchical columns containing a data frame or a matrix:
#| label: setup2
new_tbl2 <- function() {
tibble(
tb = tbl,
m = diag(4)
)
}
new_df2 <- function() {
df2 <- new_tbl2()
class(df2) <- "data.frame"
class(df2$tb) <- "data.frame"
df2
}
df2 <- new_df2()
tbl2 <- new_tbl2()
#| label: show-compare-2
#| dftbl: true
new_df()
For subset assignment (subassignment, for short), we need a fresh copy of the data for each test.
The with_*() functions (omitted here for brevity) allow for a more concise notation.
These functions take an assignment expression, execute it on a fresh copy of the data, and return the data for printing.
The first example prints what's really executed, further examples omit this output.
#| label: with-def
#| include: false
with_df <- function(code, verbose = FALSE) {
code <- rlang::enexpr(code)
full_code <- rlang::quo({
df <- new_df()
!!code
df
})
if (verbose) rlang::expr_print(rlang::quo_get_expr(full_code))
rlang::eval_tidy(full_code)
}
with_tbl <- function(code, verbose = FALSE) {
code <- rlang::enexpr(code)
full_code <- rlang::quo({
tbl <- new_tbl()
!!code
tbl
})
if (verbose) rlang::expr_print(rlang::quo_get_expr(full_code))
rlang::eval_tidy(full_code)
}
with_df2 <- function(code) {
code <- rlang::enexpr(code)
full_code <- rlang::quo({
df2 <- new_df2()
!!code
df2
})
rlang::eval_tidy(full_code)
}
with_tbl2 <- function(code) {
code <- rlang::enexpr(code)
full_code <- rlang::quo({
tbl2 <- new_tbl2()
!!code
tbl2
})
rlang::eval_tidy(full_code)
}
#| label: with-demo
#| dftbl: true
with_df(df$n <- rev(df$n), verbose = TRUE)
Column extraction
Definition of x[[j]]
x[[j]] is equal to .subset2(x, j).
#| label: double-bracket-equivalent-to-subset2
#| dftbl: true
df[[1]]
.subset2(df, 1)
#| label: double-bracket-equivalent-to-subset2-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
identical(df[[3]], .subset2(df, 3))
identical(df2[["df"]], .subset2(df2, "df"))
NB: x[[j]] always returns an object of size nrow(x) if the column exists.
#| label: double-bracket-retains-size
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
vec_size(df[[1]])
vec_size(df[[3]])
vec_size(df2[[1]])
vec_size(df2[[2]])
j must be a single number or a string, as enforced by .subset2(x, j).
#| label: double-bracket-requires-scalar-j-index
#| dftbl: true
df[[1:2]]
df[[c("n", "c")]]
df[[TRUE]]
df[[mean]]
NA indexes, numeric out-of-bounds (OOB) values, and non-integers throw an error:
#| label: double-bracket-j-oob-numeric
#| dftbl: true
df[[NA]]
df[[NA_character_]]
df[[NA_integer_]]
df[[-1]]
df[[4]]
df[[1.5]]
df[[Inf]]
Character OOB access is silent because a common package idiom is to check for the absence of a column with is.null(df[[var]]).
#| label: double-bracket-j-oob-character
#| dftbl: true
df[["x"]]
Definition of x$name
x$name and x$"name" are equal to x[["name"]].
#| label: dollar-equivalent-to-subset
#| dftbl: true
df$n
df$"n"
df[["n"]]
#| label: dollar-equivalent-to-subset-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
identical(df$li, df[["li"]])
identical(df2$tb, df2[["tb"]])
identical(df2$m, df2[["m"]])
Unlike data frames, tibbles do not partially match names.
Because df$x is rarely used in packages, it can raise a warning:
#| label: dollar-equivalent-to-subset-pmatch
#| dftbl: true
df$l
df$not_present
Column subsetting
Definition of x[j]
j is converted to an integer vector by vec_as_index(j, ncol(x), names = names(x)).
Then x[c(j_1, j_2, ..., j_n)] is equivalent to tibble(x[[j_1]], x[[j_2]], ..., x[[j_n]]), keeping the corresponding column names.
This implies that j must be a numeric or character vector, or a logical vector with length 1 or ncol(x).[^subset-extract-commute]
[^subset-extract-commute]: x[j][[jj]] is equal to x[[ j[[jj]] ]], in particular x[j][[1]] is equal to x[[j]] for scalar numeric or integer j.
#| label: bracket-j-definition
#| dftbl: true
df[1:2]
When subsetting repeated indexes, the resulting column names are undefined, do not rely on them.
#| label: bracket-j-duplication
#| dftbl: true
df[c(1, 1)]
For tibbles with repeated column names, subsetting by name uses the first matching column.
nrow(df[j]) equals nrow(df).
#| label: bracket-j-empty
#| dftbl: true
df[integer()]
Tibbles support indexing by a logical matrix, but only if all values in the returned vector are compatible.
#| label: bracket-j-logical-matrix
#| dftbl: true
df[is.na(df)]
df[!is.na(df)]
Definition of x[, j]
x[, j] is equal to x[j].
Tibbles do not perform column extraction if x[j] would yield a single column.
#| label: bracket-missing-i
#| dftbl: true
df[, 1]
df[, 1:2]
#| label: bracket-missing-i-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
identical(df[, 2:3], df[2:3])
identical(df2[, 1:2], df2[1:2])
Definition of x[, j, drop = TRUE]
For backward compatiblity, x[, j, drop = TRUE] performs column extraction, returning x[j][[1]] when ncol(x[j]) is 1.
#| label: bracket-always-returns-tibble-drop
#| dftbl: true
df[, 1, drop = TRUE]
#| label: bracket-always-returns-tibble-drop-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
identical(df[, 3, drop = TRUE], df[[3]])
identical(df2[, 1, drop = TRUE], df2[[1]])
identical(df2[, 2, drop = TRUE], df2[[2]])
Row subsetting
Definition of x[i, ]
x[i, ] is equal to tibble(vec_slice(x[[1]], i), vec_slice(x[[2]], i), ...).[^row-subset-efficiency]
[^row-subset-efficiency]: Row subsetting x[i, ] is not defined in terms of x[[j]][i] because that definition does not generalise to matrix and data frame columns.
For efficiency and backward compatibility, i is converted to an integer vector by vec_as_index(i, nrow(x)) first.
#| label: bracket-i
#| dftbl: true
df[3, ]
This means that i must be a numeric vector, or a logical vector of length nrow(x) or 1.
For compatibility, i can also be a character vector containing positive numbers.
#| label: bracket-i-wrong-type
#| dftbl: true
df[mean, ]
df[list(1), ]
df["1", ]
Exception: OOB values generate warnings instead of errors:
#| label: bracket-i-oob
#| dftbl: true
df[10, ]
df["x", ]
Unlike data frames, only logical vectors of length 1 are recycled.
#| label: bracket-i-recycle
#| dftbl: true
df[c(TRUE, FALSE), ]
NB: scalar logicals are recycled, but scalar numerics are not.
That makes the x[NA, ] and x[NA_integer_, ] return different results.
#| label: bracket-i-na
#| dftbl: true
df[NA, ]
df[NA_integer_, ]
Definition of x[i, , drop = TRUE]
drop = TRUE has no effect when not selecting a single row:
#| label: bracket-i-drop
#| dftbl: true
df[1, , drop = TRUE]
Row and column subsetting
Definition of x[] and x[,]
x[] and x[,] are equivalent to x.[^bracket-comma]
[^bracket-comma]: x[,] is equivalent to x[] because x[, j] is equivalent to x[j].
Definition of x[i, j]
x[i, j] is equal to x[i, ][j].[^bracket-flip]
[^bracket-flip]: A more efficient implementation of x[i, j] would forward to x[j][i, ].
#| label: bracket-i-j-equivalent-to-i-subset-then-j
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
df[1, 1]
df[1, ][1]
identical(df[1, 2:3], df[2:3][1, ])
identical(df[2:3, 1], df[1][2:3, ])
identical(df2[2:3, 1:2], df2[1:2][2:3, ])
Definition of x[[i, j]]
i must be a numeric vector of length 1.
x[[i, j]] is equal to x[i, ][[j]], or vctrs::vec_slice(x[[j]], i).[^bracket2-flip]
[^bracket2-flip]: Cell subsetting x[[i, j]] is not defined in terms of x[[j]][[i]] because that definition does not generalise to list, matrix and data frame columns.
A more efficient implementation of x[[i, j]] would check that j is a scalar and forward to x[i, j][[1]].
#| label: bracket-bracket-i-j-equivalent-to-i-subset-then-j
df[[1, 1]]
df[[1, 3]]
This implies that j must be a numeric or character vector of length 1.
NB: vec_size(x[[i, j]]) always equals 1.
Unlike x[i, ], x[[i, ]] is not valid.
Column update
Definition of x[[j]] <- a
If a is a vector then x[[j]] <- a replaces the jth column with value a.
#| label: double-bracket-assign-definition
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[[1]] <- 0)
with_df(df[[3]] <- 4:1)
with_df2(df2[[1]] <- 0)
with_df2(df2[[2]] <- 4:1)
#| label: double-bracket-assign-requires-scalar-j-index
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[[1]] <- 0)
with_df(df[["c"]] <- 0)
#| label: double-bracket-assign-requires-scalar-j-index-error
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[[TRUE]] <- 0)
with_df(df[[1:3]] <- 0)
with_df(df[[c("n", "c")]] <- 0)
with_df(df[[FALSE]] <- 0)
with_df(df[[1:2]] <- 0)
with_df(df[[NA_integer_]] <- 0)
with_df(df[[NA]] <- 0)
with_df(df[[NA_character_]] <- 0)
a is recycled to the same size as x so must have size nrow(x) or 1.
(The only exception is when a is NULL, as described below.)
Recycling also works for list, data frame, and matrix columns.
#| label: double-bracket-assign-recycle
#| dftbl: true
with_df(df[["li"]] <- list(0))
with_df2(df2[["tb"]] <- df[1, ])
with_df2(df2[["m"]] <- df2[["m"]][1, , drop = FALSE])
#| label: double-bracket-requires-size
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[[1]] <- 1)
with_df(df[[1]] <- 4:1)
with_df(df[[1]] <- 3:1)
with_df(df[[1]] <- 2:1)
j must be a scalar numeric or a string, and cannot be NA.
If j is OOB, a new column is added on the right hand side, with name repair if needed.
#| label: double-bracket-assign-supports-new
#| dftbl: true
with_df(df[["x"]] <- 0)
with_df(df[[4]] <- 0)
with_df(df[[5]] <- 0)
df[[j]] <- a replaces the complete column so can change the type.
#| label: double-bracket-assign-supports-type-change
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[[1]] <- df[[2]])
with_df(df[[2]] <- df[[3]])
with_df(df[[3]] <- df2[[1]])
with_df2(df2[[1]] <- df2[[2]])
with_df2(df2[[2]] <- df[[1]])
[[<- supports removing a column by assigning NULL to it.
#| label: double-bracket-assign-supports-null
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[[1]] <- NULL)
with_df2(df2[[2]] <- NULL)
Removing a nonexistent column is a no-op.
#| label: double-bracket-assign-supports-null-unknown
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[["q"]] <- NULL)
Definition of x$name <- a
x$name <- a and x$"name" <- a are equivalent to x[["name"]] <- a.[^column-assign-symmetry]
[^column-assign-symmetry]: $ behaves almost completely symmetrically to [[ when comparing subsetting and subassignment.
#| label: dollar-equivalent-to-subset-assign
#| dftbl: true
with_df(df$n <- 0)
with_df(df[["n"]] <- 0)
#| label: dollar-equivalent-to-subset-assign-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df$"n" <- 0)
$<- does not perform partial matching.
#| label: dollar-equivalent-to-subset-assign-pmatch
#| dftbl: true
with_df(df$l <- 0)
with_df(df[["l"]] <- 0)
Column subassignment: x[j] <- a
- If
j is missing, it's replaced with seq_along(x)
- If
j is logical vector, it's converted to numeric with seq_along(x)[j].
a is a list or data frame
If inherits(a, "list") or inherits(a, "data.frame") is TRUE, then x[j] <- a is equivalent to x[[j[[1]]] <- a[[1]], x[[j[[2]]]] <- a[[2]], ...
#| label: bracket-assign-def
#| dftbl: true
with_df(df[1:2] <- list("x", 4:1))
with_df(df[c("li", "x", "c")] <- list("x", 4:1, NULL))
If length(a) equals 1, then it is recycled to the same length as j.
#| label: bracket-assign-recycles
#| dftbl: true
with_df(df[1:2] <- list(1))
with_df(df[1:2] <- list(0, 0, 0))
with_df(df[1:3] <- list(0, 0))
An attempt to update the same column twice gives an error.
#| label: bracket-assign-multiple
#| dftbl: true
with_df(df[c(1, 1)] <- list(1, 2))
If a contains NULL values, the corresponding columns are removed after updating (i.e. position indexes refer to columns before any modifications).
#| label: bracket-assign-remove
#| dftbl: true
with_df(df[1:2] <- list(NULL, 4:1))
NA indexes are not supported.
#| label: bracket-assign-na
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[NA] <- list("x"))
with_df(df[NA_integer_] <- list("x"))
with_df(df[NA_character_] <- list("x"))
Just like column updates, [<- supports changing the type of an existing column.
#| label: bracket-assign-supports-type-change
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[1] <- df[2])
with_df(df[2] <- df[3])
with_df(df[3] <- df2[1])
with_df2(df2[1] <- df2[2])
with_df2(df2[2] <- df[1])
Appending columns at the end (without gaps) is supported.
The name of new columns is determined by the LHS, the RHS, or by name repair (in that order of precedence).
#| label: bracket-assign-names
#| dftbl: true
with_df(df[c("x", "y")] <- tibble("x", x = 4:1))
with_df(df[3:4] <- list("x", x = 4:1))
with_df(df[4] <- list(4:1))
with_df(df[5] <- list(4:1))
Tibbles support indexing by a logical matrix, but only for a scalar RHS, and if all columns updated are compatible with the value assigned.
#| label: bracket-j-assign-logical-matrix
#| dftbl: true
with_df(df[is.na(df)] <- 4)
with_df(df[is.na(df)] <- 1:2)
with_df(df[matrix(c(rep(TRUE, 5), rep(FALSE, 7)), ncol = 3)] <- 4)
a is a matrix or array
If is.matrix(a), then a is coerced to a data frame with as.data.frame() before assigning.
If rows are assigned, the matrix type must be compatible with all columns.
If is.array(a) and any(dim(a)[-1:-2] != 1), an error is thrown.
#| label: bracket-assign-array
#| dftbl: true
with_df(df[1:2] <- matrix(8:1, ncol = 2))
with_df(df[1:3, 1:2] <- matrix(6:1, ncol = 2))
with_df(df[1:2] <- array(4:1, dim = c(4, 1, 1)))
with_df(df[1:2] <- array(8:1, dim = c(4, 2, 1)))
with_df(df[1:2] <- array(8:1, dim = c(2, 1, 4)))
with_df(df[1:2] <- array(8:1, dim = c(4, 1, 2)))
a is another type of vector
If vec_is(a), then x[j] <- a is equivalent to x[j] <- list(a).
This is primarily provided for backward compatibility.
#| label: bracket-assign-wraps
#| dftbl: true
with_df(df[1] <- 0)
with_df(df[1] <- list(0))
Matrices must be wrapped in list() before assignment to create a matrix column.
#| label: bracket-assign-matrix
#| dftbl: true
with_df(df[1] <- list(matrix(1:8, ncol = 2)))
with_df(df[1:2] <- list(matrix(1:8, ncol = 2)))
a is NULL
Entire columns can be removed.
Specifying i is an error.
#| label: bracket-assign-null
#| dftbl: true
with_df(df[1] <- NULL)
with_df(df[, 2:3] <- NULL)
with_df(df[1, 2:3] <- NULL)
a is not a vector
Any other type for a is an error.
Note that if is.list(a) is TRUE, but inherits(a, "list") is FALSE, then a is considered to be a scalar.
See ?vec_is and ?vec_proxy for details.
#| label: bracket-assign-non-vector
#| dftbl: true
with_df(df[1] <- mean)
with_df(df[1] <- lm(mpg ~ wt, data = mtcars))
Row subassignment: x[i, ] <- list(...)
x[i, ] <- a is the same as vec_slice(x[[j_1]], i) <- a[[1]], vec_slice(x[[j_2]], i) <- a[[2]], ... .[^row-assign-symmetry]
[^row-assign-symmetry]: x[i, ] is symmetrical for subset and subassignment.
#| label: bracket-i-assign
#| dftbl: true
with_df(df[2:3, ] <- df[1, ])
with_df(df[c(FALSE, TRUE, TRUE, FALSE), ] <- df[1, ])
#| label: bracket-i-assign-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[0:2, ] <- df[1, ])
with_df(df[0, ] <- df[1, ])
with_df(df[-2, ] <- df[1, ])
with_df(df[-1:2, ] <- df[1, ])
with_df(df[NA_integer_, ] <- df[1, ])
with_df2(df2[NA_integer_, ] <- df2[1, ])
with_df(df[TRUE, ] <- df[1, ])
with_df(df[FALSE, ] <- df[1, ])
with_df(df[NA, ] <- df[1, ])
Only values of size one can be recycled.
#| label: bracket-i-recycle-assign
#| dftbl: true
with_df(df[2:3, ] <- df[1, ])
with_df(df[2:3, ] <- list(df$n[1], df$c[1:2], df$li[1]))
with_df(df[2:4, ] <- df[1:2, ])
#| label: bracket-i-recycle-assign-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df2(df2[2:4, ] <- df2[1, ])
with_df2(df2[2:4, ] <- df2[2:3, ])
For compatibility, only a warning is issued for indexing beyond the number of rows.
Appending rows right at the end of the existing data is supported, without warning.
#| label: bracket-i-oob-num
#| dftbl: true
with_df(df[5, ] <- df[1, ])
with_df(df[5:7, ] <- df[1, ])
with_df(df[6, ] <- df[1, ])
with_df(df[-5, ] <- df[1, ])
with_df(df[-(5:7), ] <- df[1, ])
with_df(df[-6, ] <- df[1, ])
For compatibility, i can also be a character vector containing positive numbers.
#| label: bracket-i-character
#| dftbl: true
with_df(df[as.character(1:3), ] <- df[1, ])
#| label: bracket-i-character-detail
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[as.character(-(1:3)), ] <- df[1, ])
with_df(df[as.character(3:5), ] <- df[1, ])
with_df(df[as.character(-(3:5)), ] <- df[1, ])
with_df(df[NA_character_, ] <- df[1, ])
Row and column subassignment
Definition of x[i, j] <- a
x[i, j] <- a is equivalent to x[i, ][j] <- a.[^bracket-assign-flip]
[^bracket-assign-flip]: x[i, j] is symmetrical for subsetting and subassignment.
A more efficient implementation of x[i, j] <- a would forward to x[j][i, ] <- a.
Subassignment to x[i, j] is stricter for tibbles than for data frames.
x[i, j] <- a can't change the data type of existing columns.
#| label: bracket-i-data-type
#| dftbl: true
with_df(df[2:3, 1] <- df[1:2, 2])
with_df(df[2:3, 2] <- df[1:2, 3])
with_df(df[2:3, 3] <- df2[1:2, 1])
with_df2(df2[2:3, 1] <- df2[1:2, 2])
with_df2(df2[2:3, 2] <- df[1:2, 1])
A notable exception is the population of a column full of NA (which is of type logical), or the use of NA on the right-hand side of the assignment.
#| label: bracket-i-j-na-init
#| dftbl: true
with_df({df$x <- NA; df[2:3, "x"] <- 3:2})
with_df({df[2:3, 2:3] <- NA})
For programming, it is always safer (and faster) to use the correct type of NA to initialize columns.
#| label: bracket-i-j-typed-na-init
#| dftbl: true
with_df({df$x <- NA_integer_; df[2:3, "x"] <- 3:2})
For new columns, x[i, j] <- a fills the unassigned rows with NA.
#| label: subassign-ij-new-column
#| dftbl: true
with_df(df[2:3, "n"] <- 1)
with_df(df[2:3, "x"] <- 1)
with_df(df[2:3, "n"] <- NULL)
Likewise, for new rows, x[i, j] <- a fills the unassigned columns with NA.
#| label: append-rows-only-all-columns
#| dftbl: true
with_df(df[5, "n"] <- list(0L))
Definition of x[[i, j]] <- a
i must be a numeric vector of length 1.
x[[i, j]] <- a is equivalent to x[i, ][[j]] <- a.[^double-bracket-ij-symmetry]
[^double-bracket-ij-symmetry]: x[[i, j]] is symmetrical for subsetting and subassignment.
An efficient implementation would check that i and j are scalar and forward to x[i, j][[1]] <- a.
#| label: double-bracket-i-j-equivalent-to-row-subset-then-j
#| dftbl: true
#| include: !expr eval_details
#| eval: !expr eval_details
with_df(df[[1, 1]] <- 0)
with_df(df[1, ][[1]] <- 0)
with_df(df[[1, 3]] <- list(NULL))
with_df(df[1, ][[3]] <- list(NULL))
with_df2(df2[[1, 1]] <- df[1, ])
with_df2(df2[1, ][[1]] <- df[1, ])
with_df2(df2[[1, 2]] <- t(1:4))
with_df2(df2[1, ][[2]] <- t(1:4))
df[[1:2, 1]]
with_df(df[[1:2, 1]] <- 0)
NB: vec_size(a) must equal 1.
Unlike x[i, ] <-, x[[i, ]] <- is not valid.
#| label: check
#| dftbl: true
#| include: false
stopifnot(identical(df, new_df()))
hadley/tibble documentation built on June 13, 2025, 3:33 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#| include: false # To suppress messages library(tibble) library(vctrs) knitr::opts_chunk$set( collapse = TRUE, comment = "#>", error = TRUE ) tibble:::set_dftbl_hooks() options( lifecycle_verbosity = "warning", lifecycle_disable_warnings = FALSE, lifecycle_verbose_soft_deprecation = TRUE, lifecycle_repeat_warnings = TRUE ) # Set to FALSE for production eval_details <- (Sys.getenv("IN_GALLEY") != "")
This vignette defines invariants for subsetting and subset-assignment for tibbles, and illustrates where their behaviour differs from data frames.
The goal is to define a small set of invariants that consistently define how behaviors interact.
Some behaviors are defined using functions of the vctrs package, e.g. vec_slice(), vec_recycle() and vec_as_index().
Refer to their documentation for more details about the invariants that they follow.
The subsetting and subassignment operators for data frames and tibbles are particularly tricky, because they support both row and column indexes, both of which are optionally missing.
We resolve this by first defining column access with [[ and $, then column-wise subsetting with [, then row-wise subsetting, then the composition of both.
Conventions
In this article, all behaviors are demonstrated using one example data frame and its tibble equivalent:
#| label: setup library(tibble) library(vctrs) new_df <- function() { df <- data.frame(n = c(1L, NA, 3L, NA)) df$c <- letters[5:8] df$li <- list(9, 10:11, 12:14, "text") df } new_tbl <- function() { as_tibble(new_df()) }
Results of the same code for data frames and tibbles are presented side by side:
#| label: show #| dftbl: true #| dftbl_always: true new_df()
If the results are identical (after converting to a data frame if necessary), only the tibble result is shown.
Subsetting operations are read-only. The same objects are reused in all examples:
#| label: ro df <- new_df() tbl <- new_tbl()
Where needed, we also show examples with hierarchical columns containing a data frame or a matrix:
#| label: setup2 new_tbl2 <- function() { tibble( tb = tbl, m = diag(4) ) } new_df2 <- function() { df2 <- new_tbl2() class(df2) <- "data.frame" class(df2$tb) <- "data.frame" df2 } df2 <- new_df2() tbl2 <- new_tbl2()
#| label: show-compare-2 #| dftbl: true new_df()
For subset assignment (subassignment, for short), we need a fresh copy of the data for each test.
The with_*() functions (omitted here for brevity) allow for a more concise notation.
These functions take an assignment expression, execute it on a fresh copy of the data, and return the data for printing.
The first example prints what's really executed, further examples omit this output.
#| label: with-def #| include: false with_df <- function(code, verbose = FALSE) { code <- rlang::enexpr(code) full_code <- rlang::quo({ df <- new_df() !!code df }) if (verbose) rlang::expr_print(rlang::quo_get_expr(full_code)) rlang::eval_tidy(full_code) } with_tbl <- function(code, verbose = FALSE) { code <- rlang::enexpr(code) full_code <- rlang::quo({ tbl <- new_tbl() !!code tbl }) if (verbose) rlang::expr_print(rlang::quo_get_expr(full_code)) rlang::eval_tidy(full_code) } with_df2 <- function(code) { code <- rlang::enexpr(code) full_code <- rlang::quo({ df2 <- new_df2() !!code df2 }) rlang::eval_tidy(full_code) } with_tbl2 <- function(code) { code <- rlang::enexpr(code) full_code <- rlang::quo({ tbl2 <- new_tbl2() !!code tbl2 }) rlang::eval_tidy(full_code) }
#| label: with-demo #| dftbl: true with_df(df$n <- rev(df$n), verbose = TRUE)
Column extraction
Definition of x[[j]]
x[[j]] is equal to .subset2(x, j).
#| label: double-bracket-equivalent-to-subset2 #| dftbl: true df[[1]] .subset2(df, 1)
#| label: double-bracket-equivalent-to-subset2-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details identical(df[[3]], .subset2(df, 3)) identical(df2[["df"]], .subset2(df2, "df"))
NB: x[[j]] always returns an object of size nrow(x) if the column exists.
#| label: double-bracket-retains-size #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details vec_size(df[[1]]) vec_size(df[[3]]) vec_size(df2[[1]]) vec_size(df2[[2]])
j must be a single number or a string, as enforced by .subset2(x, j).
#| label: double-bracket-requires-scalar-j-index #| dftbl: true df[[1:2]] df[[c("n", "c")]] df[[TRUE]] df[[mean]]
NA indexes, numeric out-of-bounds (OOB) values, and non-integers throw an error:
#| label: double-bracket-j-oob-numeric #| dftbl: true df[[NA]] df[[NA_character_]] df[[NA_integer_]] df[[-1]] df[[4]] df[[1.5]] df[[Inf]]
Character OOB access is silent because a common package idiom is to check for the absence of a column with is.null(df[[var]]).
#| label: double-bracket-j-oob-character #| dftbl: true df[["x"]]
Definition of x$name
x$name and x$"name" are equal to x[["name"]].
#| label: dollar-equivalent-to-subset #| dftbl: true df$n df$"n" df[["n"]]
#| label: dollar-equivalent-to-subset-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details identical(df$li, df[["li"]]) identical(df2$tb, df2[["tb"]]) identical(df2$m, df2[["m"]])
Unlike data frames, tibbles do not partially match names.
Because df$x is rarely used in packages, it can raise a warning:
#| label: dollar-equivalent-to-subset-pmatch #| dftbl: true df$l df$not_present
Column subsetting
Definition of x[j]
j is converted to an integer vector by vec_as_index(j, ncol(x), names = names(x)).
Then x[c(j_1, j_2, ..., j_n)] is equivalent to tibble(x[[j_1]], x[[j_2]], ..., x[[j_n]]), keeping the corresponding column names.
This implies that j must be a numeric or character vector, or a logical vector with length 1 or ncol(x).[^subset-extract-commute]
[^subset-extract-commute]: x[j][[jj]] is equal to x[[ j[[jj]] ]], in particular x[j][[1]] is equal to x[[j]] for scalar numeric or integer j.
#| label: bracket-j-definition #| dftbl: true df[1:2]
When subsetting repeated indexes, the resulting column names are undefined, do not rely on them.
#| label: bracket-j-duplication #| dftbl: true df[c(1, 1)]
For tibbles with repeated column names, subsetting by name uses the first matching column.
nrow(df[j]) equals nrow(df).
#| label: bracket-j-empty #| dftbl: true df[integer()]
Tibbles support indexing by a logical matrix, but only if all values in the returned vector are compatible.
#| label: bracket-j-logical-matrix #| dftbl: true df[is.na(df)] df[!is.na(df)]
Definition of x[, j]
x[, j] is equal to x[j].
Tibbles do not perform column extraction if x[j] would yield a single column.
#| label: bracket-missing-i #| dftbl: true df[, 1] df[, 1:2]
#| label: bracket-missing-i-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details identical(df[, 2:3], df[2:3]) identical(df2[, 1:2], df2[1:2])
Definition of x[, j, drop = TRUE]
For backward compatiblity, x[, j, drop = TRUE] performs column extraction, returning x[j][[1]] when ncol(x[j]) is 1.
#| label: bracket-always-returns-tibble-drop #| dftbl: true df[, 1, drop = TRUE]
#| label: bracket-always-returns-tibble-drop-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details identical(df[, 3, drop = TRUE], df[[3]]) identical(df2[, 1, drop = TRUE], df2[[1]]) identical(df2[, 2, drop = TRUE], df2[[2]])
Row subsetting
Definition of x[i, ]
x[i, ] is equal to tibble(vec_slice(x[[1]], i), vec_slice(x[[2]], i), ...).[^row-subset-efficiency]
[^row-subset-efficiency]: Row subsetting x[i, ] is not defined in terms of x[[j]][i] because that definition does not generalise to matrix and data frame columns.
For efficiency and backward compatibility, i is converted to an integer vector by vec_as_index(i, nrow(x)) first.
#| label: bracket-i #| dftbl: true df[3, ]
This means that i must be a numeric vector, or a logical vector of length nrow(x) or 1.
For compatibility, i can also be a character vector containing positive numbers.
#| label: bracket-i-wrong-type #| dftbl: true df[mean, ] df[list(1), ] df["1", ]
Exception: OOB values generate warnings instead of errors:
#| label: bracket-i-oob #| dftbl: true df[10, ] df["x", ]
Unlike data frames, only logical vectors of length 1 are recycled.
#| label: bracket-i-recycle #| dftbl: true df[c(TRUE, FALSE), ]
NB: scalar logicals are recycled, but scalar numerics are not.
That makes the x[NA, ] and x[NA_integer_, ] return different results.
#| label: bracket-i-na #| dftbl: true df[NA, ] df[NA_integer_, ]
Definition of x[i, , drop = TRUE]
drop = TRUE has no effect when not selecting a single row:
#| label: bracket-i-drop #| dftbl: true df[1, , drop = TRUE]
Row and column subsetting
Definition of x[] and x[,]
x[] and x[,] are equivalent to x.[^bracket-comma]
[^bracket-comma]: x[,] is equivalent to x[] because x[, j] is equivalent to x[j].
Definition of x[i, j]
x[i, j] is equal to x[i, ][j].[^bracket-flip]
[^bracket-flip]: A more efficient implementation of x[i, j] would forward to x[j][i, ].
#| label: bracket-i-j-equivalent-to-i-subset-then-j #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details df[1, 1] df[1, ][1] identical(df[1, 2:3], df[2:3][1, ]) identical(df[2:3, 1], df[1][2:3, ]) identical(df2[2:3, 1:2], df2[1:2][2:3, ])
Definition of x[[i, j]]
i must be a numeric vector of length 1.
x[[i, j]] is equal to x[i, ][[j]], or vctrs::vec_slice(x[[j]], i).[^bracket2-flip]
[^bracket2-flip]: Cell subsetting x[[i, j]] is not defined in terms of x[[j]][[i]] because that definition does not generalise to list, matrix and data frame columns.
A more efficient implementation of x[[i, j]] would check that j is a scalar and forward to x[i, j][[1]].
#| label: bracket-bracket-i-j-equivalent-to-i-subset-then-j df[[1, 1]] df[[1, 3]]
This implies that j must be a numeric or character vector of length 1.
NB: vec_size(x[[i, j]]) always equals 1.
Unlike x[i, ], x[[i, ]] is not valid.
Column update
Definition of x[[j]] <- a
If a is a vector then x[[j]] <- a replaces the jth column with value a.
#| label: double-bracket-assign-definition #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[[1]] <- 0) with_df(df[[3]] <- 4:1) with_df2(df2[[1]] <- 0) with_df2(df2[[2]] <- 4:1)
#| label: double-bracket-assign-requires-scalar-j-index #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[[1]] <- 0) with_df(df[["c"]] <- 0)
#| label: double-bracket-assign-requires-scalar-j-index-error #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[[TRUE]] <- 0) with_df(df[[1:3]] <- 0) with_df(df[[c("n", "c")]] <- 0) with_df(df[[FALSE]] <- 0) with_df(df[[1:2]] <- 0) with_df(df[[NA_integer_]] <- 0) with_df(df[[NA]] <- 0) with_df(df[[NA_character_]] <- 0)
a is recycled to the same size as x so must have size nrow(x) or 1.
(The only exception is when a is NULL, as described below.)
Recycling also works for list, data frame, and matrix columns.
#| label: double-bracket-assign-recycle #| dftbl: true with_df(df[["li"]] <- list(0)) with_df2(df2[["tb"]] <- df[1, ]) with_df2(df2[["m"]] <- df2[["m"]][1, , drop = FALSE])
#| label: double-bracket-requires-size #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[[1]] <- 1) with_df(df[[1]] <- 4:1) with_df(df[[1]] <- 3:1) with_df(df[[1]] <- 2:1)
j must be a scalar numeric or a string, and cannot be NA.
If j is OOB, a new column is added on the right hand side, with name repair if needed.
#| label: double-bracket-assign-supports-new #| dftbl: true with_df(df[["x"]] <- 0) with_df(df[[4]] <- 0) with_df(df[[5]] <- 0)
df[[j]] <- a replaces the complete column so can change the type.
#| label: double-bracket-assign-supports-type-change #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[[1]] <- df[[2]]) with_df(df[[2]] <- df[[3]]) with_df(df[[3]] <- df2[[1]]) with_df2(df2[[1]] <- df2[[2]]) with_df2(df2[[2]] <- df[[1]])
[[<- supports removing a column by assigning NULL to it.
#| label: double-bracket-assign-supports-null #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[[1]] <- NULL) with_df2(df2[[2]] <- NULL)
Removing a nonexistent column is a no-op.
#| label: double-bracket-assign-supports-null-unknown #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[["q"]] <- NULL)
Definition of x$name <- a
x$name <- a and x$"name" <- a are equivalent to x[["name"]] <- a.[^column-assign-symmetry]
[^column-assign-symmetry]: $ behaves almost completely symmetrically to [[ when comparing subsetting and subassignment.
#| label: dollar-equivalent-to-subset-assign #| dftbl: true with_df(df$n <- 0) with_df(df[["n"]] <- 0)
#| label: dollar-equivalent-to-subset-assign-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df$"n" <- 0)
$<- does not perform partial matching.
#| label: dollar-equivalent-to-subset-assign-pmatch #| dftbl: true with_df(df$l <- 0) with_df(df[["l"]] <- 0)
Column subassignment: x[j] <- a
- If
jis missing, it's replaced withseq_along(x) - If
jis logical vector, it's converted to numeric withseq_along(x)[j].
a is a list or data frame
If inherits(a, "list") or inherits(a, "data.frame") is TRUE, then x[j] <- a is equivalent to x[[j[[1]]] <- a[[1]], x[[j[[2]]]] <- a[[2]], ...
#| label: bracket-assign-def #| dftbl: true with_df(df[1:2] <- list("x", 4:1)) with_df(df[c("li", "x", "c")] <- list("x", 4:1, NULL))
If length(a) equals 1, then it is recycled to the same length as j.
#| label: bracket-assign-recycles #| dftbl: true with_df(df[1:2] <- list(1)) with_df(df[1:2] <- list(0, 0, 0)) with_df(df[1:3] <- list(0, 0))
An attempt to update the same column twice gives an error.
#| label: bracket-assign-multiple #| dftbl: true with_df(df[c(1, 1)] <- list(1, 2))
If a contains NULL values, the corresponding columns are removed after updating (i.e. position indexes refer to columns before any modifications).
#| label: bracket-assign-remove #| dftbl: true with_df(df[1:2] <- list(NULL, 4:1))
NA indexes are not supported.
#| label: bracket-assign-na #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[NA] <- list("x")) with_df(df[NA_integer_] <- list("x")) with_df(df[NA_character_] <- list("x"))
Just like column updates, [<- supports changing the type of an existing column.
#| label: bracket-assign-supports-type-change #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[1] <- df[2]) with_df(df[2] <- df[3]) with_df(df[3] <- df2[1]) with_df2(df2[1] <- df2[2]) with_df2(df2[2] <- df[1])
Appending columns at the end (without gaps) is supported. The name of new columns is determined by the LHS, the RHS, or by name repair (in that order of precedence).
#| label: bracket-assign-names #| dftbl: true with_df(df[c("x", "y")] <- tibble("x", x = 4:1)) with_df(df[3:4] <- list("x", x = 4:1)) with_df(df[4] <- list(4:1)) with_df(df[5] <- list(4:1))
Tibbles support indexing by a logical matrix, but only for a scalar RHS, and if all columns updated are compatible with the value assigned.
#| label: bracket-j-assign-logical-matrix #| dftbl: true with_df(df[is.na(df)] <- 4) with_df(df[is.na(df)] <- 1:2) with_df(df[matrix(c(rep(TRUE, 5), rep(FALSE, 7)), ncol = 3)] <- 4)
a is a matrix or array
If is.matrix(a), then a is coerced to a data frame with as.data.frame() before assigning.
If rows are assigned, the matrix type must be compatible with all columns.
If is.array(a) and any(dim(a)[-1:-2] != 1), an error is thrown.
#| label: bracket-assign-array #| dftbl: true with_df(df[1:2] <- matrix(8:1, ncol = 2)) with_df(df[1:3, 1:2] <- matrix(6:1, ncol = 2)) with_df(df[1:2] <- array(4:1, dim = c(4, 1, 1))) with_df(df[1:2] <- array(8:1, dim = c(4, 2, 1))) with_df(df[1:2] <- array(8:1, dim = c(2, 1, 4))) with_df(df[1:2] <- array(8:1, dim = c(4, 1, 2)))
a is another type of vector
If vec_is(a), then x[j] <- a is equivalent to x[j] <- list(a).
This is primarily provided for backward compatibility.
#| label: bracket-assign-wraps #| dftbl: true with_df(df[1] <- 0) with_df(df[1] <- list(0))
Matrices must be wrapped in list() before assignment to create a matrix column.
#| label: bracket-assign-matrix #| dftbl: true with_df(df[1] <- list(matrix(1:8, ncol = 2))) with_df(df[1:2] <- list(matrix(1:8, ncol = 2)))
a is NULL
Entire columns can be removed.
Specifying i is an error.
#| label: bracket-assign-null #| dftbl: true with_df(df[1] <- NULL) with_df(df[, 2:3] <- NULL) with_df(df[1, 2:3] <- NULL)
a is not a vector
Any other type for a is an error.
Note that if is.list(a) is TRUE, but inherits(a, "list") is FALSE, then a is considered to be a scalar.
See ?vec_is and ?vec_proxy for details.
#| label: bracket-assign-non-vector #| dftbl: true with_df(df[1] <- mean) with_df(df[1] <- lm(mpg ~ wt, data = mtcars))
Row subassignment: x[i, ] <- list(...)
x[i, ] <- a is the same as vec_slice(x[[j_1]], i) <- a[[1]], vec_slice(x[[j_2]], i) <- a[[2]], ... .[^row-assign-symmetry]
[^row-assign-symmetry]: x[i, ] is symmetrical for subset and subassignment.
#| label: bracket-i-assign #| dftbl: true with_df(df[2:3, ] <- df[1, ]) with_df(df[c(FALSE, TRUE, TRUE, FALSE), ] <- df[1, ])
#| label: bracket-i-assign-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[0:2, ] <- df[1, ]) with_df(df[0, ] <- df[1, ]) with_df(df[-2, ] <- df[1, ]) with_df(df[-1:2, ] <- df[1, ]) with_df(df[NA_integer_, ] <- df[1, ]) with_df2(df2[NA_integer_, ] <- df2[1, ]) with_df(df[TRUE, ] <- df[1, ]) with_df(df[FALSE, ] <- df[1, ]) with_df(df[NA, ] <- df[1, ])
Only values of size one can be recycled.
#| label: bracket-i-recycle-assign #| dftbl: true with_df(df[2:3, ] <- df[1, ]) with_df(df[2:3, ] <- list(df$n[1], df$c[1:2], df$li[1])) with_df(df[2:4, ] <- df[1:2, ])
#| label: bracket-i-recycle-assign-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df2(df2[2:4, ] <- df2[1, ]) with_df2(df2[2:4, ] <- df2[2:3, ])
For compatibility, only a warning is issued for indexing beyond the number of rows. Appending rows right at the end of the existing data is supported, without warning.
#| label: bracket-i-oob-num #| dftbl: true with_df(df[5, ] <- df[1, ]) with_df(df[5:7, ] <- df[1, ]) with_df(df[6, ] <- df[1, ]) with_df(df[-5, ] <- df[1, ]) with_df(df[-(5:7), ] <- df[1, ]) with_df(df[-6, ] <- df[1, ])
For compatibility, i can also be a character vector containing positive numbers.
#| label: bracket-i-character #| dftbl: true with_df(df[as.character(1:3), ] <- df[1, ])
#| label: bracket-i-character-detail #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[as.character(-(1:3)), ] <- df[1, ]) with_df(df[as.character(3:5), ] <- df[1, ]) with_df(df[as.character(-(3:5)), ] <- df[1, ]) with_df(df[NA_character_, ] <- df[1, ])
Row and column subassignment
Definition of x[i, j] <- a
x[i, j] <- a is equivalent to x[i, ][j] <- a.[^bracket-assign-flip]
[^bracket-assign-flip]: x[i, j] is symmetrical for subsetting and subassignment.
A more efficient implementation of x[i, j] <- a would forward to x[j][i, ] <- a.
Subassignment to x[i, j] is stricter for tibbles than for data frames.
x[i, j] <- a can't change the data type of existing columns.
#| label: bracket-i-data-type #| dftbl: true with_df(df[2:3, 1] <- df[1:2, 2]) with_df(df[2:3, 2] <- df[1:2, 3]) with_df(df[2:3, 3] <- df2[1:2, 1]) with_df2(df2[2:3, 1] <- df2[1:2, 2]) with_df2(df2[2:3, 2] <- df[1:2, 1])
A notable exception is the population of a column full of NA (which is of type logical), or the use of NA on the right-hand side of the assignment.
#| label: bracket-i-j-na-init #| dftbl: true with_df({df$x <- NA; df[2:3, "x"] <- 3:2}) with_df({df[2:3, 2:3] <- NA})
For programming, it is always safer (and faster) to use the correct type of NA to initialize columns.
#| label: bracket-i-j-typed-na-init #| dftbl: true with_df({df$x <- NA_integer_; df[2:3, "x"] <- 3:2})
For new columns, x[i, j] <- a fills the unassigned rows with NA.
#| label: subassign-ij-new-column #| dftbl: true with_df(df[2:3, "n"] <- 1) with_df(df[2:3, "x"] <- 1) with_df(df[2:3, "n"] <- NULL)
Likewise, for new rows, x[i, j] <- a fills the unassigned columns with NA.
#| label: append-rows-only-all-columns #| dftbl: true with_df(df[5, "n"] <- list(0L))
Definition of x[[i, j]] <- a
i must be a numeric vector of length 1.
x[[i, j]] <- a is equivalent to x[i, ][[j]] <- a.[^double-bracket-ij-symmetry]
[^double-bracket-ij-symmetry]: x[[i, j]] is symmetrical for subsetting and subassignment.
An efficient implementation would check that i and j are scalar and forward to x[i, j][[1]] <- a.
#| label: double-bracket-i-j-equivalent-to-row-subset-then-j #| dftbl: true #| include: !expr eval_details #| eval: !expr eval_details with_df(df[[1, 1]] <- 0) with_df(df[1, ][[1]] <- 0) with_df(df[[1, 3]] <- list(NULL)) with_df(df[1, ][[3]] <- list(NULL)) with_df2(df2[[1, 1]] <- df[1, ]) with_df2(df2[1, ][[1]] <- df[1, ]) with_df2(df2[[1, 2]] <- t(1:4)) with_df2(df2[1, ][[2]] <- t(1:4)) df[[1:2, 1]] with_df(df[[1:2, 1]] <- 0)
NB: vec_size(a) must equal 1.
Unlike x[i, ] <-, x[[i, ]] <- is not valid.
#| label: check #| dftbl: true #| include: false stopifnot(identical(df, new_df()))
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.