Remove variables from survey data (step)

Creates a step that removes one or more variables from the survey data when baked.

Usage

step_remove(
  svy,
  ...,
  vars = NULL,
  .copy = use_copy_default(),
  comment = "Remove variables",
  use_copy = deprecated(),
  lazy = lazy_default(),
  record = TRUE
)

Arguments

svy

A Survey or RotativePanelSurvey object

...

Unquoted variable names to remove, or a character vector

vars

Character vector of variable names to remove. Alternative to ... for programmatic use.

.copy

Whether to operate on a copy (default: use_copy_default())

comment

Descriptive text for the step for documentation and traceability (default "Remove variables").

use_copy

Use .copy instead.

lazy

Internal. Whether to delay execution (default lazy_default()).

record

Internal. Whether to record the step (default TRUE).

Value

Survey object with the specified variables removed (or queued for removal).

Details

Lazy evaluation (default): By default, steps are recorded but not executed until bake_steps() is called.

Variables can be specified in two ways:

• Unquoted names: step_remove(svy, age, income)

• Character vector: step_remove(svy, vars = c("age", "income"))

Name resolution: bare symbols passed via ... that match a column name always refer to that column, even if a variable with the same name exists in the calling environment (column takes precedence). Symbols that do not match any column are looked up in the calling environment and, if they hold a character vector, its values are used as column names. For fully programmatic use prefer vars = c("age", "income"), which skips symbol resolution entirely.

Variables that don't exist in the data produce a warning (not an error), allowing pipelines to be robust to missing columns.

See also

Other steps: bake_steps(), get_steps(), step_collapse(), step_compute(), step_filter(), step_join(), step_quantile(), step_recode(), step_rename(), step_validate(), view_graph()

Examples

dt <- data.table::data.table(
  id = 1:5, age = c(25, 30, 45, 50, 60),
  w = rep(1, 5)
)
svy <- Survey$new(
  data = dt, edition = "2023", type = "ech",
  psu = NULL, engine = "data.table", weight = add_weight(annual = "w")
)
svy2 <- step_remove(svy, age)
svy2 <- bake_steps(svy2)
"age" %in% names(get_data(svy2)) # FALSE
#> [1] FALSE