Skip to content

56 Allow custom importance scaling#57

Merged
jfkcooper merged 12 commits intomainfrom
56---Allow-custom-importance-scaling
Jul 28, 2023
Merged

56 Allow custom importance scaling#57
jfkcooper merged 12 commits intomainfrom
56---Allow-custom-importance-scaling

Conversation

@sstendahl
Copy link
Collaborator

Closes issue #56.

Adds an optional argument to the fisher function, allowing to scale the importance of each parameter with a predefined value, on top of the regular unit scaling. As per relevant paper

@sstendahl sstendahl linked an issue Jul 24, 2023 that may be closed by this pull request
@sstendahl sstendahl requested a review from jfkcooper July 27, 2023 10:40
@sstendahl
Copy link
Collaborator Author

Should be OK now, the changes outside of the fisher function happened when I gave Pycharm permission to solve the PEP8 conflicts. They did stop my linter from complaining.

Copy link
Owner

@jfkcooper jfkcooper left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sorry, a couple more comments!

@sstendahl sstendahl requested a review from jfkcooper July 27, 2023 15:55
Copy link
Owner

@jfkcooper jfkcooper left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good!

@jfkcooper jfkcooper merged commit e38d4e6 into main Jul 28, 2023
@jfkcooper jfkcooper deleted the 56---Allow-custom-importance-scaling branch July 28, 2023 08:49
sstendahl pushed a commit to sstendahl/HOGBEN that referenced this pull request Sep 6, 2023
…rtance-scaling

56  Allow custom importance scaling
sstendahl pushed a commit to sstendahl/HOGBEN that referenced this pull request Sep 6, 2023
…rtance-scaling

56  Allow custom importance scaling
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Allow users to add importance scaling for the parameters explicitly

2 participants