Interpolate short missing gaps in Gazepoint pupil data

Performs linear interpolation over short internal gaps in Gazepoint pupil data. This function is intended to be used after flag_gazepoint_pupil() or flag_gazepoint_pupil_artifacts(). When available, pupil_clean is used as the preferred default input column, followed by pupil_for_preprocessing. Leading gaps, trailing gaps, long gaps, non-finite time values, and groups with too few valid pupil samples are not interpolated.

Usage

interpolate_gazepoint_pupil(
  data,
  pupil_col = NULL,
  time_col = NULL,
  group_cols = c("subject", "media_id"),
  max_gap_ms = 150,
  max_gap_samples = Inf,
  min_valid_points = 2
)

Arguments

data

A Gazepoint master table, preferably after flag_gazepoint_pupil() or flag_gazepoint_pupil_artifacts().

pupil_col

Optional name of the pupil column to interpolate. If NULL, the function detects one of pupil_clean, pupil_for_preprocessing, mean_pupil, pupil, pupil_raw, left_pupil, or right_pupil.

time_col

Optional name of the time column. If NULL, the function detects one of time_ms, time, time_orig, or time_orig_ms.

group_cols

Character vector of grouping columns used to keep interpolation within independent time series. Defaults to c("subject", "media_id") using internally standardised names. Use character(0) for global interpolation.

max_gap_ms

Maximum duration, in milliseconds, of a gap that may be interpolated. The duration is measured between the valid samples immediately before and after the gap. Defaults to 150.

max_gap_samples

Maximum number of consecutive missing samples that may be interpolated. Defaults to Inf.

min_valid_points

Minimum number of valid samples required within a group before interpolation is attempted. Defaults to 2.

Value

A tibble containing the original data plus interpolation columns.

Examples

if (FALSE) { # \dontrun{
flagged <- flag_gazepoint_pupil(master)

interpolated <- interpolate_gazepoint_pupil(flagged)

dplyr::count(interpolated, pupil_interpolation_status)

artifact_flagged <- flag_gazepoint_pupil_artifacts(master)

artifact_interpolated <- interpolate_gazepoint_pupil(artifact_flagged)

dplyr::count(artifact_interpolated, pupil_interpolation_status)
} # }