Why does plot behave differently for different data, and how do you control it? My specific example today is getting different output with scaled and non-scaled data. For example
fit <- kmeans(mydata, 4)
plot(mydata, col = fit$cluster)
returns a nice 4x4 scatter matrix plot (mydata is 486x4) but
mydata <- scale(mydata)
fit <- kmeans(mydata, 4)
plot(mydata, col = fit$cluster)
returns a single plot showing variable 1 scatter plotted vs variable 2?