Bind multiple data frames by row or column

Bind multiple data frames by row or column. These functions are similar to do.call("rbind", list) or do.call("cbind", list). See details for further usage.

Usage

bind_rows(x)

bind_fill_rows(x)

bind_cols(x)

Arguments

x: A list of data frames.

Value

data.frame

Details

bind_rows() is used for a list of data frames that have the same column names. bind_fill_rows() is used for a list of data frames that have different column names (filling in missing rows with NA). See below for how each handles column types:

Function	Factor	List	Date	POSIXct	POSIXlt
`bind_rows`	Supported	Supported	Supported	Supported	Supported
`bind_fill_rows`	Coerced to integer	Supported	Coerced to integer	Coerced to integer	Error

bind_rows() is faster than do.call("rbind", list), but much slower than dplyr::bind_rows() and data.table::rbindlist() for large lists. bind_fill_rows() is slower.

bind_cols() produces the same results as do.call("cbind", list), but requires that each data frame have the same number of rows.

Before use, ensure data types are compatible per column and that each column has a proper name. bind_rows() will convert factors to character if columns of the same name have mixed character/factor types. If the same column names have different factor levels or level ordering then test output before using.

Examples

#----------------------------------------------------------------------------
# bind() examples
#----------------------------------------------------------------------------
library(bkdat)

l1 <- list(
  data.frame(a = c(1, 2)),
  data.frame(a = c(3, 4))
)
l2 <- list(
  data.frame(a = c(1, 2)),
  data.frame(b = c(3, 4))
)
l3 <- list(
  as_df(
    list(
      POSIXct = as.POSIXct(c("2017-08-13", "2017-08-14")),
      POSIXlt = as.POSIXlt(c("2017-08-13", "2017-08-14")),
      Date = as.Date(c("2017-08-13", "2017-08-14")),
      numeric = c(1, 2),
      integer = 3:4,
      character = letters[1:2],
      missing = rep(c(1, NA)),
      logical = rep(c(TRUE, FALSE)),
      factor = factor(letters[1:2]),
      list = replicate(2, list(a=data.frame(b = c(1, 2))))
    )
  ),
  as_df(
    list(
      POSIXct = as.POSIXct(c("2017-08-13", "2017-08-14")),
      POSIXlt = as.POSIXlt(c("2017-08-13", "2017-08-14")),
      Date = as.Date(c("2017-08-13", "2017-08-14")),
      numeric = c(1, 2),
      integer = 3:4,
      character = letters[1:2],
      missing = rep(c(1, NA)),
      logical = rep(c(TRUE, FALSE)),
      factor = factor(letters[1:2]),
      list = replicate(2, list(a=data.frame(b = c(1, 2))))
    )
  )
)

bind_rows(l1)
#>   a
#> 1 1
#> 2 2
#> 3 3
#> 4 4
bind_fill_rows(l2)
#>    a  b
#> 1  1 NA
#> 2  2 NA
#> 3 NA  3
#> 4 NA  4
bind_rows(l3)
#>      POSIXct    POSIXlt       Date numeric integer character missing logical
#> 1 2017-08-13 2017-08-13 2017-08-13       1       3         a       1    TRUE
#> 2 2017-08-14 2017-08-14 2017-08-14       2       4         b      NA   FALSE
#> 3 2017-08-13 2017-08-13 2017-08-13       1       3         a       1    TRUE
#> 4 2017-08-14 2017-08-14 2017-08-14       2       4         b      NA   FALSE
#>   factor list
#> 1      a 1, 2
#> 2      b 1, 2
#> 3      a 1, 2
#> 4      b 1, 2