Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Potential problem when ntree << N #6

Closed
fradav opened this issue May 10, 2019 · 1 comment
Closed

Potential problem when ntree << N #6

fradav opened this issue May 10, 2019 · 1 comment
Assignees
@fradav
Labels
bug Something isn't working model-choice related to the model choice methodology

Comments

@fradav
Copy link
Collaborator

fradav commented May 10, 2019 •
edited
Loading

When the number of trees is small wrt the number of samples, there is an increased risk of having some samples not in any OOB of any trees. We can't get OOB prediction for those and they should simply removed in the training set for the error regression part.

For the moment they are just "NAN" and "crash" the error regression training, which gives 0 as posterior probability.
@fradav fradav added bug Something isn't working model-choice related to the model choice methodology labels May 10, 2019
@fradav fradav self-assigned this May 10, 2019
@fradav fradav changed the title Potential crash when ntree << N Potential problem when ntree << N May 10, 2019
@fradav
Copy link
Collaborator Author

fradav commented May 10, 2019

Fixed in #6726e9c
We can now use the methods with only one tree, for example.

@fradav fradav closed this as completed May 10, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@fradav fradav

Labels
bug Something isn't working model-choice related to the model choice methodology
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@fradav