Skip to contents

Align Multiple Channels to a Common Grid

Source: R/step_channel_align.R
step_measure_channel_align.Rd

step_measure_channel_align() creates a specification of a recipe step that aligns multiple measurement channels to a common location grid.

Usage

step_measure_channel_align(
  recipe,
  ...,
  method = c("union", "intersection", "reference"),
  reference = 1L,
  interpolation = c("linear", "spline", "constant"),
  role = NA,
  trained = FALSE,
  skip = FALSE,
  id = recipes::rand_id("measure_channel_align")
)

Arguments

recipe

A recipe object.

...

One or more selector functions to choose measure columns. If empty, all measure columns are used.

method

How to determine the common grid:

  • "union" (default): Use all unique locations from all channels

  • "intersection": Use only locations present in all channels

  • "reference": Use the grid from the reference channel

reference

For method = "reference", which channel to use as reference. Can be a column name (character) or column index (integer). Default is 1 (first channel).

interpolation

Interpolation method for missing values:

  • "linear" (default): Linear interpolation

  • "spline": Cubic spline interpolation

  • "constant": Nearest neighbor (constant)

role

Not used.

trained

Logical indicating if the step has been trained.

skip

Logical. Should the step be skipped when baking?

id

Unique step identifier.

Value

An updated recipe with the new step added.

Details

Multi-channel analytical instruments (e.g., LC-DAD, SEC with multiple detectors) often produce measurements at slightly different location grids for each channel. This step aligns all channels to a common grid, enabling:

  • Direct comparison between channels

  • Channel combination or ratio calculations

  • Modeling with consistent feature dimensions

Grid Methods

  • Union: Creates a grid containing all unique locations from all channels. Values are interpolated where channels don't have data.

  • Intersection: Uses only locations where all channels have data. No interpolation needed but may lose data at edges.

  • Reference: Uses one channel's grid as the target. Other channels are interpolated to match.

See also

Other measure-channel: step_measure_channel_combine(), step_measure_channel_ratio()

Examples

library(recipes)
library(tibble)

# Create sample multi-channel data
df <- tibble(
  id = rep(1:3, each = 10),
  time_uv = rep(seq(0, 9, by = 1), 3),
  absorbance_uv = rnorm(30, 100, 10),
  time_ri = rep(seq(0.5, 9.5, by = 1), 3),
  absorbance_ri = rnorm(30, 50, 5),
  concentration = rep(c(10, 25, 50), each = 10)
)

# Ingest as separate channels, then align
rec <- recipe(concentration ~ ., data = df) |>
  update_role(id, new_role = "id") |>
  step_measure_input_long(absorbance_uv, location = vars(time_uv)) |>
  step_measure_input_long(absorbance_ri, location = vars(time_ri)) |>
  step_measure_channel_align(method = "union")