This function performs k-means clustering on spatial or non-spatial data and determines the optimal number of clusters based on either the within-cluster sum of squares or the average silhouette width.
Usage
optimal_kmeans_cluster(
data,
spatial = TRUE,
coords = NULL,
max_cluster = 2L,
method = c("wss", "silhouette")
)Arguments
- data
The input data for clustering. It can be an 'sf' object, a 'SpatialPoints' object, or a data frame with coordinates.
- spatial
Logical indicating whether the input data is spatial (default is TRUE). If set to FALSE, the data is assumed to be non-spatial and the clustering is performed on the provided coordinates.
- coords
The column names of the coordinates in the data (required if spatial is set to FALSE).
- max_cluster
The maximum number of clusters to consider.
- method
The method to determine the optimal number of clusters. It can be "wss" (within-cluster sum of squares) or "silhouette" (average silhouette width).
Value
A list containing the clustering results (data frame) and the plot of the selected method (ggplot object).
Examples
if (FALSE) {
data("landcover")
okc <- optimal_kmeans_cluster(data = landcover,spatial = TRUE,coords = NULL,
max_cluster = 15, method = "wss")
# look at data
okc$data
}