Skip to contents

Transpose a data.frame or tibble

Source: R/tidy_t.R
tidy_t.Rd

Allow transposing a data.frame or tibble in a "tidy" way. Instead of row names, the column names of the transposed object are taken from a specified id_col.
Note: The input df should not contain row names, as they would be discarded.

Usage

tidy_t(
  df,
  id_col = 1,
  exclude_cols = NULL,
  new_id_colname = NULL,
  reguess_coltypes = FALSE,
  name_repair = "unique",
  verbose = FALSE
)

Arguments

df

The to-be-transposed data.frame or tibble.

id_col

Which column of df shall be used as new column names? Can be either an index or a column name.

exclude_cols

Exclude specified columns before transposing. Can be either one or multiple indices or a column names, or NULL.

new_id_colname

Rename the first column of the transposed object. If NULL, carry over the selected id_col column name.

reguess_coltypes

Shall a re-guessing of column types be performed after transposing? Allows to correctly retrieve e.g. numeric or logical columns by utilizing reguess_coltypes().

name_repair

Choose strategy to potentially repair 'id_col', e.g. for uniqueness. Is passed on to vctrs::vec_as_names(), supporting among else the following strategies:

  • "minimal": No name repair or checks, beyond basic existence of names and replacing NA's.

  • "unique": Make sure names are unique and not empty. A suffix is appended to duplicate names to make them unique.

  • "universal": Make the names unique and syntactic, meaning that you can safely use the names as variables without causing a syntax error.

verbose

Set verbosity of potential name_repair.

Value

A transposed object - keeps object type of df input (data.frame or tibble).

Examples

df <- data.frame(a = c("x1", "1", "FALSE"), b = c("x2", "2", "TRUE"))
df
#>       a    b
#> 1    x1   x2
#> 2     1    2
#> 3 FALSE TRUE

tidy_t(df)
#>   a x1 1 FALSE
#> 1 b x2 2  TRUE

# reguessing column types
tidy_t(df, reguess_coltypes = TRUE)
#>   a x1 1 FALSE
#> 1 b x2 2  TRUE

# using second column as id column
tidy_t(df, id_col = "b")
#>   b x2 2  TRUE
#> 1 a x1 1 FALSE

# excluding columns
tidy_t(df, exclude_cols = "b")
#> [1] a     x1    1     FALSE
#> <0 rows> (or 0-length row.names)

# renaming id column
tidy_t(df, new_id_colname = "var")
#>   var x1 1 FALSE
#> 1   b x2 2  TRUE