I am using superpc package to carry out supervised principal component analysis on my data. I am using the following link as reference
http://statweb.stanford.edu/~tibs/superpc/tutorial
I am using regression analysis for my data.
When I plot the CVPlot, I get the following plot
As per my understanding the likelihood statistic ratio should be highest for 3rd analysis and so should be the threshold value. However I am getting an irregular plot. Can someone guide if I have interpreted it correctly or not. And due to what reasons can the graph be irregular?