Skip to contents

Calculate Signal-to-Noise Ratio

Source: R/qc.R
step_measure_qc_snr.Rd

step_measure_qc_snr() creates a specification of a recipe step that calculates the signal-to-noise ratio (SNR) for each measurement and adds it as a new column. This is useful for quality control and filtering.

Usage

step_measure_qc_snr(
  recipe,
  measures = NULL,
  new_col = ".snr",
  signal_method = c("max", "range", "rms"),
  noise_method = c("diff", "mad", "residual"),
  role = "predictor",
  trained = FALSE,
  skip = FALSE,
  id = recipes::rand_id("measure_qc_snr")
)

Arguments

recipe

A recipe object.

measures

An optional character vector of measure column names.

new_col

Name of the new column to store SNR values. Default is ".snr".

signal_method

How to estimate signal:

  • "max" (default): Maximum absolute value

  • "range": Peak-to-peak range

  • "rms": Root mean square

noise_method

How to estimate noise:

  • "diff" (default): RMS of first differences (estimates high-freq noise)

  • "mad": Median absolute deviation of values

  • "residual": Residuals from smoothed fit

role

Role for the new column. Default is "predictor".

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

SNR is calculated as signal / noise, where signal and noise are estimated using the specified methods. Higher values indicate cleaner data.

The "diff" noise method is particularly useful because it estimates high-frequency noise without being affected by broad spectral features: $$noise = \sqrt{\frac{1}{2(n-1)} \sum_{i=2}^{n} (x_i - x_{i-1})^2}$$

See also

Other measure-qc: step_measure_impute(), step_measure_qc_outlier(), step_measure_qc_saturated()

Examples

library(recipes)

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

bake(rec, new_data = NULL)
#> # A tibble: 215 × 7
#>       id water   fat protein .measures channel      .snr
#>    <int> <dbl> <dbl>   <dbl>    <meas> <list>      <dbl>
#>  1     1  60.5  22.5    16.7 [100 × 2] <int [100]>  277.
#>  2     2  46    40.1    13.5 [100 × 2] <int [100]>  370.
#>  3     3  71     8.4    20.5 [100 × 2] <int [100]>  291.
#>  4     4  72.8   5.9    20.7 [100 × 2] <int [100]>  278.
#>  5     5  58.3  25.5    15.5 [100 × 2] <int [100]>  271.
#>  6     6  44    42.7    13.7 [100 × 2] <int [100]>  374.
#>  7     7  44    42.7    13.7 [100 × 2] <int [100]>  374.
#>  8     8  69.3  10.6    19.3 [100 × 2] <int [100]>  255.
#>  9     9  61.4  19.9    17.7 [100 × 2] <int [100]>  310.
#> 10    10  61.4  19.9    17.7 [100 × 2] <int [100]>  303.
#> # ℹ 205 more rows