I am not sure about some parts of the following plots:

-- I do not understand what does the expression on x-axis means (first figure) and how to interpret that.

-- What is the bars and the points in figure 2 and 3 refer to.

Could you please help me ?


enter image description here

enter image description here

enter image description here

  • $\begingroup$ It's impossible to answer this question without knowing what software and commands you used to create the plots. $\endgroup$ – Sycorax says Reinstate Monica Aug 12 '16 at 19:34
  • $\begingroup$ I used "lars" function that available in R program, I just would like to understand it generally. what do they mean ? $\endgroup$ – Ereck Aug 12 '16 at 19:35
  • $\begingroup$ Because I'm sure you read the documentation before posting this question, what part of the lars documentation is unclear to you? $\endgroup$ – Sycorax says Reinstate Monica Aug 12 '16 at 19:35
  • $\begingroup$ @ General Abrial, Of course, but I don,t know about the expression on x-axis (figure 1) $\endgroup$ – Ereck Aug 12 '16 at 19:38
  • 2
    $\begingroup$ The lars function isn't in base R, so it must be in some package, but you haven't specified one. Please add a reproducible example for people to work with. Then this will be answerable. $\endgroup$ – gung - Reinstate Monica Aug 12 '16 at 20:07
  1. The x-axis is the size of the coefficient relative to the $L_1$ norm of the coefficient vector.... probably. I'm assuming that you created that plot with the default plot commands in lars and this is what the documentation describes.

  2. The bars in figures 2 and 3 represent an estimate of the sampling error around the mean estimate of MSE. I don't know how you generated them, so I don't know how wide the interval is. Default values are described in the documentation. The red points in figure 3 are the mean estimates of the MSE. The colored vertical lines are the locations of the minimum MSE and minimum MSE plus 1 standard error (by default).

  • $\begingroup$ 1- so, is this expression related to the value of the penalty term $\lambda$ ,and what is max $\beta$ here. 2- some people said the red point is MSE for resulting model and other said is cross-validation curve. are they equivalent to the mean estimates , also mean estimates of what? Thanks $\endgroup$ – Ereck Aug 12 '16 at 20:28
  • 1
    $\begingroup$ @Ereck 1. yes 2. They said the red point was MSE because it is. Look at the label for the Y axis of the plot. Then consider what the meaning of the initials "MSE." $\endgroup$ – Sycorax says Reinstate Monica Aug 12 '16 at 20:31
  • $\begingroup$ @ General Abrial, I know MSE= mean square error. 1- in which sense they are related. $\endgroup$ – Ereck Aug 12 '16 at 20:38
  • $\begingroup$ are the points in figure 2 also refer to the mean estimates of the MSE? $\endgroup$ – Ereck Aug 12 '16 at 21:04
  • $\begingroup$ as far as I can tell $\endgroup$ – Sycorax says Reinstate Monica Aug 12 '16 at 21:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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