Extract representative scanpaths

Select representative observed scanpaths by minimizing mean distance to other members of the same fitted cluster.

Usage

extract_gazepoint_representative_scanpaths(x, n_per_cluster = 1L)

Arguments

x

An object returned by cluster_gazepoint_scanpaths().

n_per_cluster

Number of representatives to return per cluster.

Value

A data frame containing cluster, representative rank, sequence identifier, mean within-cluster distance, cluster size, and PAM medoid status.

Examples

d <- matrix(
  c(
    0, 0.1, 1, 1,
    0.1, 0, 1, 1,
    1, 1, 0, 0.1,
    1, 1, 0.1, 0
  ),
  nrow = 4,
  byrow = TRUE,
  dimnames = list(LETTERS[1:4], LETTERS[1:4])
)

fit <- cluster_gazepoint_scanpaths(d, k = 2)
extract_gazepoint_representative_scanpaths(fit)
#>   cluster representative_rank sequence_id mean_within_cluster_distance
#> 1       1                   1           A                          0.1
#> 2       2                   1           C                          0.1
#>   cluster_size is_model_medoid
#> 1            2           FALSE
#> 2            2           FALSE