Compute cluster-wise local outlier factor scores with dbscan::lof()

Uses dbscan::lof() to compute Local Outlier Factor (LOF) scores separately within each cluster of a ctdf, and writes the result into the lof column.

Usage

ctdf_lof(ctdf, minPts = NULL)

Arguments

ctdf

A ctdf object .

minPts

Optional integer passed to dbscan::lof(). If NULL, dbscan::lof() is called with its defaults.

The function updates ctdf by reference by updating the lof column.

See also

dbscan::lof(), cluster_track().

Examples

data(mini_ruff)
x = as_ctdf(mini_ruff)
cluster_track(x)
#> → Find putative cluster regions.
#> ! Repairing[1]...
#> → Local clustering.
#> ! Repairing[2]...
#> ! Compute lof scores...
x = ctdf_lof(x)
head(x[, .(.id, cluster, lof)])
#> Key: <.id>
#>      .id cluster      lof
#>    <int>   <int>    <num>
#> 1:     1       0       NA
#> 2:     2       0       NA
#> 3:     3       0       NA
#> 4:     4       0       NA
#> 5:     5       1 1.315277
#> 6:     6       1 1.184660