Calculate CHAOS score

CHAOS score measures the clustering performance by calculating the mean length of the graph edges in the 1-nearest neighbor (1NN) graph for each cluster, averaged across clusters. Lower CHAOS score indicates better spatial domain clustering performance.

Usage

CHAOS(labels, location, BNPARAM = NULL)

Arguments

labels

Cluster labels.

location

A numeric data matrix containing location information, where rows are points and columns are location dimensions.

BNPARAM

BNPARAM object passed to findKNN specifying the kNN approximation method to use. Defaults to exact for small datasets, and Annoy for larger ones.

Value

A numeric value for CHAOS score.

Examples

data(sp_toys)
data <- sp_toys
CHAOS(data$label, data[,c("x", "y")])
#> $CHAOS
#> [1] 0.2305866
#> 
#> $CHAOS_class
#>         2         1 
#> 0.2305866 0.2305866 
#> 
CHAOS(data$p1, data[,c("x", "y")])
#> $CHAOS
#> [1] 0.2305866
#> 
#> $CHAOS_class
#>         2         1 
#> 0.2305866 0.2305866 
#> 
CHAOS(data$p2, data[,c("x", "y")])
#> $CHAOS
#> [1] 0.2305866
#> 
#> $CHAOS_class
#>         2         1 
#> 0.2305866 0.2305866 
#>