2
$\begingroup$

The image shown below is the output for plot(lm.fit) on the Boston dataset, where lm.fit=lm(medv~lstat) . Why are some of the points labelled in the plot ? enter image description here

$\endgroup$
1
  • $\begingroup$ This is purely a matter of reading the documentation. The help page for plot.lm, which is the function that produced these plots, states "id.n: number of points to be labelled in each plot, starting with the most extreme" is the argument controlling how many points will be labeled. By default this value is 3 (as the source code attests). The source code also shows that "extreme" means "largest absolute residuals." $\endgroup$ – whuber May 19 '17 at 13:45
1
$\begingroup$

The are considered as point of interest, e.g., outlier, high leverage, high residual points, for user to do investigation.

See this post for details on how to read the plots.

Interpreting plot.lm()

$\endgroup$
1
  • $\begingroup$ This does not correctly describe the plots shown in the question. From the bottom right plot it is clear that the points are not the highest leverage ones. Only the points with largest absolute residuals are labeled. $\endgroup$ – whuber May 19 '17 at 13:47

Not the answer you're looking for? Browse other questions tagged or ask your own question.