weights_graph {cvxclustr} | R Documentation |
Constructs the adjacency matrix of the weights graph. This is useful to determine the connectivity of the weights graph.
weights_graph(w, n)
w |
Weights vector |
n |
Number of points being clustered |
## Clusterpaths for Mammal Dentition data(mammals) X <- as.matrix(mammals[,-1]) X <- t(scale(X,center=TRUE,scale=FALSE)) n <- ncol(X) ## Pick some weights and a sequence of regularization parameters. k <- 5 phi <- 0.5 w <- kernel_weights(X,phi) w <- knn_weights(w,k,n) A <- weights_graph(w,n) find_clusters(A) ## Visualize Cluster Assignment G <- graph.adjacency(A, mode = 'upper') plot(G,vertex.label=as.character(mammals[,1]),vertex.label.cex=0.65,vertex.label.font=2)