Dear people,
I used clusplot to plot a partition result. The partition result was from
pamk with a distance object as input. Then I applied cmdscale on the same
distance object for coordinates to make another scatterplot.
My problem is this: the coordinates from the cmdscale calculation, though
with the same shape, were different in scale and rotation from the scatter
plot yielded by clusplot. clusplot is scaled around [-20, 30], while
cmdscale is scaled between [-1,1]. Also in my case, the two plots are
mirroring of each other. Is there a way to replicate the way that clusplot
makes use of cmdscale? Or just yielding more comparable plots through the
two functions?
I know both clusplot and cmdscale decided the rotation based on princomp.
Then where come these differences? I tried to look at the code through
typing "clusplot" in the console but it was brief and the complex methods
seem to hide somewhere else.
Thank you in advance for your kind help.
Kind regards,
Junru
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
