TODO

 - [ ] Replace AVL code
   - https://github.com/ebiggers/avl_tree
   - https://github.com/greensky00/avltree
   - https://github.com/willemt/array-avl-tree
 - [ ] Merge eaf/eaf3d/uc/trunk/README
 - [ ] Merge eaf/eaf3d/uc/trunk/ command-line improvements
 - [ ] Merge hypervolume/Makefile README
 - [ ] Coverage Error Sayin 2024: is this the same as R2 indicator?
   - Sayın, S.: Measuring the quality of discrete representations of efficient sets in multiple objective mathematical programming. Math. Program. 87(3), 543–560 (2000)
   - Sayın, S.: A procedure to find discrete representations of the efficient set with specified coverage errors. Oper. Res. 51, 427–436 (2003
 - [ ] Appendix B of "The Hypervolume Indicator: Computational Problems and Algorithms" has more ND sets types: wave, cliff and hard.
 - [ ] Document logHYP: T. Friedrich, K. Bringmann, T. Voß, C. Igel. The Logarithmic Hypervolume Indicator. 11th ACM Foundations of Genetic Algorithms (FOGA), pages 81-92, 2009. https://hpi.de/friedrich/publications/all/Document/puma-friedrich/FriedrichBVI11.pdf/a839ddd47688611dcf92512ca010daf3.html?tx_extbibsonomycsl_publicationlist%5Baction%5D=view&tx_extbibsonomycsl_publicationlist%5Bpreview%5D=large&cHash=7cabd8bc9699c0ab3f88886a3f62976f
 - [ ] Document relation between hypervolume and attainment function.
 - [ ] R2 indicator: [Reinvestigating the R2 Indicator: Achieving Pareto Compliance by Integration](https://link.springer.com/chapter/10.1007/978-3-031-70085-9_13) [Brockhoff, D., Wagner, T., Trautmann, H.: On the properties of the R2 indicator. In: Proceedings of the 14th Annual Conference on Genetic and Evolutionary
Computation, pp. 465–472 (2012)]
 - [ ] archiving: https://github.com/CMA-ES/moarchiving
 - [ ] Convert code from https://github.com/fieldsend/multiobjective_data_structures
 - [ ] Take a look at https://github.com/esa/pagmo2/blob/master/src/utils/multi_objective.cpp
 - [ ] Faster epsilon: https://github.com/mbuzdalov/orthant-search/blob/master/implementations/src/main/java/ru/ifmo/orthant/epsilon/OrthantImplementation.java
       and https://dl.acm.org/doi/pdf/10.1145/3205455.3205469 and http://www.cmap.polytechnique.fr/~nikolaus.hansen/proceedings/2016/GECCO/proceedings/p613.pdf
 - Online nondominated filter
   - [ ] [A Fast Incremental BSP Tree Archive for Non-dominated Points](https://doi.org/10.1007/978-3-319-54157-0_18)
 - [ ] Faster nondominated algorithm (is_nondominated, filter_dominated) for very many points: [Worst-case I/O-efficient Skyline Algorithms](https://www.cse.cuhk.edu.hk/~taoyf/paper/tods12-sky.pdf)
 - [ ] Faster nondominated sorting:
    - Jensen improvement: [Fortin et al. 2013](https://dl.acm.org/doi/epdf/10.1145/2463372.2463454)
      [DEAP](https://github.com/DEAP/deap/blob/1dc8d28ec06d44cda01bfa86c1c02b49fcc1969d/deap/tools/emo.py#L240)
    - https://github.com/mbuzdalov/non-dominated-sorting       https://github.com/KernelA/nds-py
 <https://discourse.julialang.org/t/fast-non-dominated-sorting/24164/20>
 <https://ctlab.itmo.ru/~mbuzdalov/ec-2018/05-nds.pdf>
    - Tree partitioning by Gustavson/Syberfeldt
    - [Merge non-dominated sorting](https://github.com/jMetal/jMetal/blob/main/jmetal-core/src/main/java/org/uma/jmetal/util/ranking/impl/MergeNonDominatedSortRanking.java
)
    - Jensen-style divide-and-conquer with small-size subroutines?
 - [ ] 3D Nondominated sorting:
     Y. Nekrich. A fast algorithm for three-dimensional layers of maxima problem. In Algorithms and Data Structures, number 6844 in Lecture Notes in Computer Science, pages 607–618. 2011. [Paper](https://link.springer.com/chapter/10.1007/978-3-642-22300-6_51)
 - [ ] Epsilon-dominance:
   - https://github.com/matthewjwoodruff/pareto.py
 - [ ] HSSP (Hypervolume Subset Selection Problem)
   - https://github.com/rgoomes/hssp
 - [ ] Exactly computing Expected Hypervolume Improvement (EHVI) and qEHVI for any dimension: https://github.com/Ksrma/EHVI-HVI
 - [ ] Decremental greedy hypervolume subset selection: https://github.com/Ksrma/gHSSD-Tree
 - [ ] Real-Time Hypervolume Estimation with Monotonically Reducing Error: https://github.com/fieldsend/hypervolume
 - [ ] Benchmark against: https://github.com/MOEAFramework/MOEAFramework
 - [ ] Setup https://github.com/koaning/sphinx-marimo
 - [x] Merge hypervolume/trunk README
 - [x] hv2d contributions
 - [X] pre-commit for R: https://lorenzwalthert.github.io/precommit/articles/ci.html
 - [X] create moocore R package
 - [X] create moocore Python package
 - [X] Setup github actions
 - [X] Rewrite regtest.pl in python.
 - [X] move eaftools/eaf code here.
 - [x] move hypervolume/trunk code here.
 - [x] move hypervolume testsuite here.
 - [x] Faster nondominated algorithm (is_nondominated, filter_dominated) for d>3:
    - Kung et al.: https://doi.org/10.1145/321906.321910
    - [Bentley: Multidimensional Divide-and-Conquer](https://dl.acm.org/doi/pdf/10.1145/358841.358850)
    - [C++ implementation](https://github.com/TLDart/nondLib/blob/main/nondlib.hpp) and [paper about the implementation](https://dl.acm.org/doi/10.1145/3449726.3462737)