[PRE REVIEW]: Balsa: A Fast C++ Random Forest Classifier with Commandline and Python Interface

[editorialbot]

Submitting author: @tobiasborsdorff (Tobias Borsdorff)
Repository: https://github.com/SRON-Earth/Balsa
Branch with paper.md (empty if default branch):
Version: v1.0.0
Editor: @HaoZeke
Reviewers: Pending
Managing EiC: Chris Vernon

Status

Status badge code:

HTML: <a href="https://joss.theoj.org/papers/f324b8495db8e2983e97cb9692817b48"><img src="https://joss.theoj.org/papers/f324b8495db8e2983e97cb9692817b48/status.svg"></a>
Markdown: [![status](https://joss.theoj.org/papers/f324b8495db8e2983e97cb9692817b48/status.svg)](https://joss.theoj.org/papers/f324b8495db8e2983e97cb9692817b48)

Author instructions

Thanks for submitting your paper to JOSS @tobiasborsdorff. Currently, there isn't a JOSS editor assigned to your paper.

@tobiasborsdorff if you have any suggestions for potential reviewers then please mention them here in this thread (without tagging them with an @). You can search the list of people that have already agreed to review and may be suitable for this submission.

Editor instructions

The JOSS submission bot @editorialbot is here to help you find and assign reviewers and start the main review. To find out what @editorialbot can do for you type:

@editorialbot commands

[editorialbot]

Hello human, I'm @editorialbot, a robot that can help you with some common editorial tasks.

For a list of things I can do to help you, just type:

@editorialbot commands

For example, to regenerate the paper pdf after making changes in the paper's md or bib files, type:

@editorialbot generate pdf

[editorialbot]

Software report:

github.com/AlDanial/cloc v 1.90  T=0.03 s (1658.5 files/s, 282255.4 lines/s)
-------------------------------------------------------------------------------
Language                     files          blank        comment           code
-------------------------------------------------------------------------------
C/C++ Header                    24            702           1133           2515
C++                             16            347            231           1854
Markdown                         2            343              0            725
TeX                              1             12              0             82
Bourne Shell                     1             13             15             70
YAML                             1              0              0             55
CMake                            3             20              7             45
-------------------------------------------------------------------------------
SUM:                            48           1437           1386           5346
-------------------------------------------------------------------------------

Commit count by author:

   136	Joris van Zwieten
   104	Denis de Leeuw Duarte
     5	Tobias Borsdorff

[editorialbot]

Reference check summary (note 'MISSING' DOIs are suggestions that need verification):

✅ OK DOIs

- 10.3390/rs16071208 is OK
- 10.5281/zenodo.14186320 is OK
- 10.5281/zenodo.14186406 is OK
- 10.1023/A:1010933404324 is OK
- 10.5194/amt-14-665-2021 is OK
- 10.5194/amt-16-1597-2023 is OK

🟡 SKIP DOIs

- No DOI given, and none found for title: Balsa: A Fast C++ Random Forest Classifier
- No DOI given, and none found for title: Scikit-learn: Machine Learning in Python

❌ MISSING DOIs

- None

❌ INVALID DOIs

- None

[editorialbot]

Paper file info:

📄 Wordcount for paper.md is 861

✅ The paper includes a Statement of need section

[editorialbot]

License info:

✅ License found: BSD 3-Clause "New" or "Revised" License (Valid open source OSI approved license)

[editorialbot]

👉📄 Download article proof 📄 View article proof on GitHub 📄 👈

[editorialbot]

Five most similar historical JOSS papers:

ADaPT-ML: A Data Programming Template for Machine Learning
Submitting author: @nulberry
Handling editor: @jmschrei (Active)
Reviewers: @aaronpeikert, @wincowgerDEV
Similarity score: 0.7037

ASCENDS: Advanced data SCiENce toolkit for Non-Data Scientists
Submitting author: @ornlpmcp
Handling editor: @terrytangyuan (Retired)
Reviewers: @zhampel, @jrbourbeau
Similarity score: 0.6984

AutoClassWrapper: a Python wrapper for AutoClass C classification
Submitting author: @pierrepo
Handling editor: @trallard (Retired)
Reviewers: @rpetit3, @lowandrew
Similarity score: 0.6955

rFBP: Replicated Focusing Belief Propagation algorithm
Submitting author: @Nico-Curti
Handling editor: @arokem (Retired)
Reviewers: @justusschock, @DanielLenz
Similarity score: 0.6913

CRATE: A Python package to perform fast material simulations
Submitting author: @BernardoFerreira
Handling editor: @Kevin-Mattheus-Moerman (Active)
Reviewers: @RahulSundar, @atzberg, @Extraweich, @kingyin3613
Similarity score: 0.6906

⚠️ Note to editors: If these papers look like they might be a good match, click through to the review issue for that paper and invite one or more of the authors before considering asking the reviewers of these papers to review again for JOSS.

[crvernon]

@editorialbot invite @HaoZeke as editor

👋 can you take this one on @HaoZeke?

[editorialbot]

Invitation to edit this submission sent!

[tobiasborsdorff]

@editorialbot invite @HaoZeke as editor

👋 can you take this one on @HaoZeke?

Dear @crvernon,

I hope this message finds you well. I’m writing to kindly follow up on the paper I submitted over a month ago. It seems that the editor assignment has not yet been initiated. Could you please let me know when this step might take place?

I truly appreciate your time and assistance and look forward to your response.

Best regards, Tobias Borsdorff

[crvernon]

👋 @tobiasborsdorff - we have a large backlog of submissions right now so it may take a little longer than usual to get you set up with an editor. Thanks for your patience!

@HaoZeke - are you able to take this one on?

[crvernon]

@HaoZeke just checking back in on this one.

[HaoZeke]

@editorialbot assign @HaoZeke as editor

Thanks for the invite @crvernon

[editorialbot]

Assigned! @HaoZeke is now the editor

[HaoZeke]

Hi @dostuffthatmatters 👋 would you be interested in and available to review this JOSS submission? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: joss.readthedocs.io/en/latest/review_criteria.html

[HaoZeke]

Hi @cpellet 👋 would you be interested in and available to review this JOSS submission? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: joss.readthedocs.io/en/latest/review_criteria.html

[HaoZeke]

Hi @bcjaeger 👋 would you be interested in and available to review this JOSS submission? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: joss.readthedocs.io/en/latest/review_criteria.html

[dostuffthatmatters]

Hi @HaoZeke,

Thank you for asking. I am busy with another review right now, so I have to kindly decline.

A general remark:

The main statement of need for this tool is about performance. But there is no performance evaluation against existing tools in the paper. The ScikitLearn implementations for decision trees and random forests are extremely efficient and there are many ML libraries like ScikitLearn with interfaces for different programming languages.

Maybe there is a good reason for the existence of Balsa, but I could not tell that from the given material. Good luck with the review!

[bcjaeger]

Hi @bcjaeger 👋 would you be interested in and available to review this JOSS submission? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: joss.readthedocs.io/en/latest/review_criteria.html

Hello! 👋 This looks very interesting, but I don't have enough availability to review at the moment. I agree that pointing to a formal benchmark of computational efficiency in the article would be a great addition.

[tobiasborsdorff]

Dear @HaoZeke,

I hope you're doing well. I wanted to follow up regarding the review process for my manuscript, as it has not yet started. If possible, could you kindly check whether potential reviewers are available?

I appreciate your time and assistance.

Best regards,
Tobias Borsdorff

[HaoZeke]

My apologies @tobiasborsdorff, however, have there been any updates addressing the comments of @dostuffthatmatters ?

[tobiasborsdorff]

Dear @HaoZeke,

Thank you for your feedback. We have already conducted a performance analysis of the Balsa algorithm, which is documented in Section 6.1 of the ATBD https://zenodo.org/records/14186320 that is also referenced in our paper.

As illustrated by the figures in that chapter, Balsa demonstrates clear advantages in both memory usage and runtime compared to the scikit-learn implementation as well as ranger—a C++ implementation of the random forest algorithm.

We would be happy to include a brief summary of these results in the manuscript to enhance clarity. However, we kindly ask that this addition be addressed during the official review process. While @dostuffthatmatters raised a valuable point, it is important to note that the suggestion was made outside the formal review framework and the manuscript is still not under review.

Best regards, Tobias Borsdorff

[tobiasborsdorff]

Dear @HaoZeke and @dostuffthatmatters,

I hope this message finds you well. I wanted to inform you that I have updated the manuscript to include a full paragraph including figures discussing the runtime and memory usage of the Balsa implementation in comparison with the SKlearn - Python and Ranger C++ implementations. This addition highlights the performance advantages of Balsa, particularly in terms of memory efficiency and prediction speed making the software suitability for large-scale operations.

With this update, I believe the manuscript is now complete and ready for the next steps. As the manuscript has been awaiting review since December last year, I would greatly appreciate it if you could kindly restart the review process and initiate the search for reviewers at your earliest convenience.

Thank you for your time and consideration.

Best regards, Tobias

[HaoZeke]

Thanks @tobiasborsdorff , if you have any suggested reviewers please let me know without the @ here.

[dostuffthatmatters]

@editorialbot generate pdf

[editorialbot]

👉📄 Download article proof 📄 View article proof on GitHub 📄 👈

[editorialbot]

Five most similar historical JOSS papers:

None
Submitting author: None
Handling editor: None (None)
Reviewers: None
Similarity score: 0.6816

None
Submitting author: None
Handling editor: None (None)
Reviewers: None
Similarity score: 0.6762

MetObs - a Python toolkit for using non-traditional meteorological observations
Submitting author: @vergauwenthomas
Handling editor: @hugoledoux (Active)
Reviewers: @ashwinvis, @Zeitsperre
Similarity score: 0.6761

quantile-forest: A Python Package for Quantile Regression Forests
Submitting author: @reidjohnson
Handling editor: @jbytecode (Active)
Reviewers: @jncraton, @oparisot
Similarity score: 0.6750

rFBP: Replicated Focusing Belief Propagation algorithm
Submitting author: @Nico-Curti
Handling editor: @arokem (Retired)
Reviewers: @justusschock, @DanielLenz
Similarity score: 0.6710

⚠️ Note to editors: If these papers look like they might be a good match, click through to the review issue for that paper and invite one or more of the authors before considering asking the reviewers of these papers to review again for JOSS.

[dostuffthatmatters]

@tobiasborsdorff great, thanks - looks promising! My other review is still ongoing though. Best of luck!

[tobiasborsdorff]

Thanks @tobiasborsdorff , if you have any suggested reviewers please let me know without the @ here.

Dear @HaoZeke, I don’t know many people who review for JOSS, but a colleague of mine suggested these two: https://github.com/mkhorton and https://github.com/dgasmith. Maybe they’re a good fit! It’d be great if you could also reach out to a few people you think might be interested. regards Tobias