Interpolate Gazepoint pupil data using PCHIP — interpolate_gazepoint_pupil

Apply shape-preserving piecewise cubic Hermite interpolation to short internal gaps in Gazepoint pupil time series. This helper is intended as a sensitivity branch alongside the default linear interpolation workflow. It does not overwrite the original pupil column.

Usage

interpolate_gazepoint_pupil_pchip(
  data,
  pupil_col = NULL,
  time_col = NULL,
  grouping_cols = NULL,
  max_gap_ms = 500,
  max_gap_samples = NULL,
  min_valid_points = 3,
  output_col = "pupil_interpolated_pchip",
  flag_col = "interpolated_pupil_pchip",
  status_col = "pchip_interpolation_status"
)

Arguments

data: A data frame containing pupil time-series data.
pupil_col: Name of the pupil column to interpolate. If NULL, common processed pupil columns are detected automatically.
time_col: Name of the time column. If NULL, common time columns are detected automatically.
grouping_cols: Optional character vector of grouping columns. If NULL, common participant/trial/media grouping columns are detected automatically. Use character(0) to interpolate the full data as one sequence.
max_gap_ms: Maximum internal gap duration, in milliseconds, eligible for interpolation. If both max_gap_ms and max_gap_samples are supplied, both criteria must be satisfied.
max_gap_samples: Maximum number of consecutive missing samples eligible for interpolation.
min_valid_points: Minimum number of valid non-missing points required within a group before PCHIP interpolation is attempted.
output_col: Name of the interpolated pupil output column.
flag_col: Name of the logical flag column indicating samples filled by PCHIP interpolation.
status_col: Name of the interpolation-status column.

Value

A tibble with PCHIP interpolation columns added.