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

permutation importance throws warnings #110

Closed
pplonski opened this issue Jul 7, 2020 · 1 comment
Closed

permutation importance throws warnings #110

pplonski opened this issue Jul 7, 2020 · 1 comment
Assignees
@pplonski
Labels
bug Something isn't working
Milestone
version 0.5.0

Comments

@pplonski
Copy link
Contributor

pplonski commented Jul 7, 2020

when training neural network on Audult dataset I have warnings:

divide by zero encountered in log
invalid value encountered in multiply
@pplonski pplonski self-assigned this Jul 9, 2020
@pplonski pplonski added the bug Something isn't working label Jul 9, 2020
@pplonski pplonski added this to the version 0.4.2 milestone Jul 9, 2020
@pplonski
Copy link
Contributor Author

pplonski commented Jul 9, 2020

Need to set eps=1e-7 in log_loss in permutation importance scorer. By default, it assumes that predictions will be in float64 and from NN they were in float32.

@pplonski pplonski closed this as completed Jul 9, 2020
pplonski added a commit that referenced this issue Jul 9, 2020
@pplonski
set logloss eps in permutation importance (#110)
710d16c
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@pplonski pplonski

Labels
bug Something isn't working
Projects
None yet
Milestone
version 0.5.0
Development

No branches or pull requests
1 participant
@pplonski