Flag MAD-based EDA wearable artifacts — flag_gazepoint_mad

Flags dependency-light, subject-specific EDA artifact categories using robust median absolute deviation (MAD) logic. The categories are heuristic QC labels: step artifacts, needle artifacts, flatline artifacts, and wall artifacts.

Usage

flag_gazepoint_mad_artifacts(
  dat,
  eda_col = "GSR_US",
  time_col = NULL,
  group_cols = NULL,
  mad_multiplier = 8,
  flatline_tolerance = 1e-06,
  flatline_min_run = 5,
  wall_abs_change = NULL,
  output_prefix = "mad"
)

Arguments

dat: A data frame.
eda_col: Numeric EDA/conductance column.
time_col: Optional time column for ordering within group.
group_cols: Optional grouping columns.
mad_multiplier: MAD multiplier used for robust thresholding.
flatline_tolerance: Maximum absolute sample-to-sample change treated as flatline.
flatline_min_run: Minimum consecutive flatline samples.
wall_abs_change: Optional absolute change threshold for wall artifacts.
output_prefix: Prefix for output columns.

Value

A data frame with artifact flags and artifact-summary attributes.