My residuals vs. leverage plot:
shows a legend for Cook's distance lines but doesn't actually show any, though it does highlight three potentially influential points.
I tried very simply plot(leave_MRM1,5)
, so I was expecting to see them. This is in a Jupyter R Notebook.
max(cooks.distance(leave_MRM1))
?lm
diagnostic plots have somehow been flagged as potentially influential points; in fact they are simply the three most extreme points (presumably those with the largest Cook's distance, even if this is not particularly large) in each plot: from ?plot.lm
:id.n [default = 3]: number of points to be labelled in each plot, starting with the most extreme.