Apply a Function to Each Sample's Measurements

measure_map() applies a function to each sample's measurement data. This function is intended for exploration and prototyping, not for production pipelines. For reproducible preprocessing, use step_measure_map() instead.

Usage

measure_map(
  .data,
  .f,
  .cols = NULL,
  ...,
  verbosity = 1L,
  .error_call = rlang::caller_env()
)

Arguments

.data

A data frame containing one or more measure_list columns.

.f

A function or formula to apply to each sample's measurement tibble.

• If a function, it is used as-is.

• If a formula (e.g., ~ { .x$value <- log(.x$value); .x }), it is converted to a function using rlang::as_function().

.cols

<tidy-select> Columns to apply the transformation to. Defaults to all measure_list columns.

...

Additional arguments passed to .f.

verbosity

An integer controlling output verbosity:

• 0: Silent - suppress all messages and output from .f

• 1: Normal (default) - show output from .f

.error_call

The execution environment for error reporting.

Value

A data frame with the specified measure columns transformed.

Details

Intended Use: Exploration, Not Production

This function is designed for interactive exploration and debugging:

# Good: Prototyping a new transformation
baked_data |>
  measure_map(~ { .x$value <- my_experimental_fn(.x$value); .x })

# Better: Once it works, put it in a recipe step
recipe(...) |>
  step_measure_map(my_experimental_fn) |>
  prep()

Unlike recipe steps, transformations applied with measure_map() are NOT:

• Automatically applied to new data

• Bundled into workflows

• Reproducible across sessions

Function Requirements

The function .f must:

• Accept a tibble with location and value columns

• Return a tibble with location and value columns

• Not change the number of rows

See also

• step_measure_map() for production use in recipe pipelines

• measure_map_safely() for fault-tolerant exploration

• measure_summarize() for computing summary statistics

Examples

library(recipes)

# First, get data in internal format
rec <- recipe(water + fat + protein ~ ., data = meats_long) |>
  update_role(id, new_role = "id") |>
  step_measure_input_long(transmittance, location = vars(channel)) |>
  prep()

baked_data <- bake(rec, new_data = NULL)

# Explore a custom transformation
result <- measure_map(baked_data, ~ {
  # Subtract the minimum value from each spectrum
  .x$value <- .x$value - min(.x$value)
  .x
})

# Once you're happy with it, use step_measure_map() in your recipe:
# recipe(...) |>
#   step_measure_map(~ { .x$value <- .x$value - min(.x$value); .x })