DBSCAN core / border / noise points (Geron Ch 8, morie.fn grdbs)

The eps-neighbourhood INCLUDES the point itself, so min_samples = 1 makes everything core. neighbors entries are 0-based index vectors.

Usage

morie_geron_dbscan_core_point(X, eps, min_samples, metric = "euclidean")

Arguments

X

Points (m, d).

eps

Positive radius.

min_samples

Density threshold, >= 1.

metric

"euclidean", "manhattan" or "chebyshev".

Value

List with is_core, is_border, is_noise, neighbor_counts, neighbors, n_core, n_noise.

Examples

morie_geron_dbscan_core_point(X = c(1, 2, 3, 4, 5, 6, 7, 8), eps = 0.5, min_samples = 5L)
#> $is_core
#> [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> 
#> $is_border
#> [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> 
#> $is_noise
#> [1] TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE
#> 
#> $neighbor_counts
#> [1] 1 1 1 1 1 1 1 1
#> 
#> $neighbors
#> $neighbors[[1]]
#> [1] 0
#> 
#> $neighbors[[2]]
#> [1] 1
#> 
#> $neighbors[[3]]
#> [1] 2
#> 
#> $neighbors[[4]]
#> [1] 3
#> 
#> $neighbors[[5]]
#> [1] 4
#> 
#> $neighbors[[6]]
#> [1] 5
#> 
#> $neighbors[[7]]
#> [1] 6
#> 
#> $neighbors[[8]]
#> [1] 7
#> 
#> 
#> $n_core
#> [1] 0
#> 
#> $n_noise
#> [1] 8
#> 
#> $eps
#> [1] 0.5
#> 
#> $min_samples
#> [1] 5
#> 
#> $estimate
#> [1] 0
#> 
#> $n
#> [1] 8
#> 
#> $method
#> [1] "DBSCAN core-point identification"
#>