Skip to contents

Compute AOI scanpath similarity

Source: R/scanpath_similarity.R
compute_gazepoint_scanpath_similarity.Rd

Compute pairwise AOI-sequence similarity between grouped scanpaths using a lightweight Levenshtein edit-distance implementation. Similarity is reported as 1 - normalized_distance, where normalized distance is divided by the longer sequence length.

Usage

compute_gazepoint_scanpath_similarity(
  data,
  aoi_col,
  group_cols,
  time_col = NULL,
  include_missing = FALSE,
  missing_label = "missing",
  collapse_repeats = FALSE,
  max_sequences = 200
)

Arguments

data

A data frame containing AOI observations.

aoi_col

Name of the AOI column.

group_cols

Columns defining each scanpath, for example subject and trial.

time_col

Optional time/order column.

include_missing

Should missing AOI labels be retained as a state?

missing_label

Label used when retaining missing AOIs.

collapse_repeats

Should consecutive repeated AOI labels be collapsed?

max_sequences

Maximum number of grouped sequences to compare.

Value

A long-format data frame containing pairwise edit distances, normalized distances, and similarities.