TMVA regression output code simplification #3694
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Next commit will modify some files in TMVA. This commit formats the files to be changed in the following commit.
The EvaluateAllMethods in Factory.cxx is a monster. This commit starts a simplification for the regression methods. Mainly, the many vectors of vectors can be replaces with a single vector of a struct. The same struct can be used for Training/Testing, therefore reducing code duplication. The "UsefulSort" function (a bubble sort, used in combination with a lot of copies of vectors) isn't needed any longer and can be replaced with a simple std::sort. Furthermore, the methods will now be sorted correctly. Before this commit, the methods were sorted by absolute deviation even though the label said "sorted by RMS". This caused a "sorted mostly correct" effect. To allow this, the 10-argument (9 out, 1 in) void function 'TestRegression' has been changed to only one argument. It now returns a struct of the 9 former outout values instead.
Simplifies AdaBoostR2 code, avoids "UsefulSort..." function.
|
Can one of the admins verify this patch? |
Contributor
|
Thank you very much for the PR! Sorry for the very late reply 🙁 The PR also has merge conflicts by now. TMVA (all parts except for SOFIE) is in maintenance mode, so I would refrain from large scale refactoring at this point. Thank you for your understanding, and let me know if I miss some context about why this refactor is beneficial for ROOT users! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Please have a look at this simplification commit.
I am not 100% happy with the new names I have introduced.
A similiar commit should be made for Classification as well. The EvaluateAllMethods functions is still a monster.
As for the modification in the AdaBoost: that is the only untested change in this commit (please test the change).
For easier reviewing, I have separated this pull in 4 commits. The clang-format just changes whitespace. The main commit is "[TMVA] Simplify code for regression summary".
Why this commit?
##############
Reason was the wrong order of results. There were not ordered by RMS. Looking at the code revealed the mess :-)