将点添加到散点图矩阵

Question

I want to add points to my Scatter plot matrix.

head(pca)
        pc1       pc2         pc3       pc4       
[1,]  0.72859559 -2.2864943 -0.5408501  0.1564730  
[2,]  0.34852943  0.3100891  0.6007349 -0.5985266  
[3,] -0.04605026  0.5067896 -0.2911211 -1.1617171  
[4,] -1.88358617  1.3739440 -0.5655383  0.9518367  
[5,]  0.35528650 -1.7482304 -0.3871520 -0.7837712  

I ran kmeans to my PCA data and I can plot a scatter plot matrix by using the following:

k <- kmeans(pca,3)

plot(pca, col=k$clust, pch=16)

Next I want to add X's to show the centers by adding

+ points(k$centers, col=1:3 ,pch="X", cex=3)

but its doesn't work. If I reduce to only 2 PCs then the following plots the centers but only for 2 pcs:

plot(pca[,c(1,2)], col=k$clust, pch=16) + points(k$centers, col=1:10 ,pch="X", cex=3)

side note when I enter the code above R returns "numeric(0)". What does that mean?

  > plot(pca[,c(1,2)], col=k$clust, pch=16) + points(k$centers, col=1:10 ,pch="X", cex=3)
  > numeric(0)

the following has been helpful to some degree. Kmeans clustering identifying knowledge in R

Answer 1

I'm not sure what you mean by "It's not working". This works fine for me:

# parameters
rows <- 100
cols <- 4
groups <- 3

# create a matrix of uniforms
pca <- matrix(runif(rows*cols),ncol = cols,byrow=TRUE)
colnames(pca) <- paste0('pc',seq(cols))

# calculate the kmeans
(k <- kmeans(pca,groups))

# plot the data and the means
plot(pca, col=k$clust, pch=15)
points(k$centers, col=seq(groups),pch="X", cex=3)

note that you're only going to get a voronoi tessellation in a 2-d plot when the number of columns ( cols ) in your matrix is 2.