Closed GaryBAYLOR closed 2 years ago
I'll also post related comments under this issue, to keep things organized.
contributions: the contribution link https://github.com/blind-contours/CVtreeMLE/blob/master/CONTRIBUTING.md
does not lead anywhere
I'll add here the following:
I have updated the brief explanation of the simulation and the respective "mixture cube". I've increased the nfolds to 5, originally it was smaller for easier testing of the package. I've added a real world example in the vignette and there is a breast cancer example in the tests. I've included discussion of ML via SL3 in the readme and vignette.
Simulation plots In the plot that explains the simulation settings, please explain what does the plot on the left (with orange cube) represent (co-variables? then why three dimensions?), what is the relationship between A, W and M1, M2, M3? What is the difference between M1, M2,M3 in the left plot with those in the right plot?
number of folds in cross-validation It makes sense to use a higher value of
n_folds
instead of 2 to see how the results look like, as typically in cross-validation, higher number of folds(e.g.,n_folds
= 4, 8, 10, etc.) are used.