Skip to contents

Setup DBSCANConfig

Source: R/120_ClusteringConfig.R
setup_DBSCAN.Rd

Setup DBSCANConfig

Usage

setup_DBSCAN(
  eps = 0.5,
  min_points = 5L,
  weights = NULL,
  border_points = TRUE,
  search = "kdtree",
  bucket_size = 100L,
  split_rule = "SUGGEST",
  approx = 0,
  features = NULL
)

Arguments

eps

Numeric (0, Inf): Radius of neighborhood.

min_points

Integer [1, Inf): Minimum number of points in a neighborhood to form a cluster.

weights

Optional Numeric vector: Weights for data points.

border_points

Logical: If TRUE, assign border points to clusters.

Character {"kdtree", "linear", "dist"}: Nearest neighbor search strategy.

bucket_size

Integer [1, Inf): Size of buckets for k-d tree search.

split_rule

Character {"SUGGEST", "STD", "MIDPT", "FAIR", "SL_MIDPT", "SL_FAIR"}: Rule for splitting the k-d tree.

approx

Numeric [0, Inf): Approximation factor for the nearest neighbor search (dbscan::frNN(approx =)); 0 searches exactly. A positive value guarantees only the neighbors within eps / (1 + approx) and may omit others within eps, which can produce spurious clusters and noise points in exchange for speed.

features

Optional Character vector: Names of at least 2 distinct feature columns to cluster on. NULL clusters on all numeric features.

Value

DBSCANConfig object.

Author

EDG

Examples

dbscan_config <- setup_DBSCAN(eps = 0.5, min_points = 5L)
dbscan_config
#> <DBSCAN ClusteringConfig>
#>           eps: <nmr> 0.50
#>    min_points: <int> 5
#>       weights: <NUL> NULL
#> border_points: <lgc> TRUE
#>        search: <chr> kdtree
#>   bucket_size: <int> 100
#>    split_rule: <chr> SUGGEST
#>        approx: <nmr> 0.00