Journal article · 2026 · Open access

qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization

S. Ashwin Renganathan, Peter E. Bachman, Kade Carlson

The Journal of Open Source Software, 11(125), 10867 · Published 22 September 2026 · DOI 10.21105/joss.10867

Open PDFView at the publisherCite this paper

In brief

“qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization” is a 2026 journal article by S. Ashwin Renganathan and 2 co-authors, published in The Journal of Open Source Software. OpenAlex counts 0 citations as of 19 November 2026. It is open access under a CC BY licence: the full text is free to read at The Journal of Open Source Software.

Citations
0
References
30
Authors
3

Abstract

Designing an engineered system involves optimizing the system for some performance objective under several design constraints.In practice, though, there is typically more than one objective to meet and the objectives may conflict with each other.For instance, an aircraft designer may need to reduce the aerodynamic drag on the aircraft while ensuring passenger safety.Aerodynamic efficiency typically leads to slender aircraft cross-sections which can, however, compromise the safety of the aircraft under gusty conditions.Similar tradeoffs occur in chemical process optimization, materials discovery, and machine learning hyperparameter optimization.Optimizing simultaneously for more than one objective is called multiobjective optimization, and the solution to such problems is no longer a single point but a set of points that offer a pragmatic compromise between the objectives.qPOTS is a Python package that solves constrained multiobjective optimization when there is no derivative information and every query of the objectives and constraints is expensive, thus necessitating "sample efficiency" (i.e., solving the problem with as few evaluations of objectives and constraints as possible).The package implements batch Pareto optimal Thompson sampling (qPOTS), a Bayesian optimization (BO) method (Frazier, 2018;Garnett, 2023;Renganathan & Carlson, 2025).The method uses Thompson sampling (Thompson, 1933), and the prefix 𝑞 denotes the number of design points selected in each batch.The package also includes qPOTS-Decoupled, which supports decoupled objective and constraint evaluations, that is, when objectives and constraints can be evaluated independently by computer models using separate resources.In this setting, the algorithm improves sample efficiency by evaluating only a subset of objectives and constraints during each design iteration (Renganathan et al., 2026).The cited manuscript calls this method qPOTS-DOE, where DOE stands for decoupled oracle evaluations.qPOTS is released under the GNU General Public License v3.0 and is built on widely used scientific Python libraries including PyTorch (Paszke et al., 2019), GPyTorch (Gardner et al., 2018), BoTorch (Balandat et al., 2020), and Pymoo (Blank & Deb, 2020).

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Renganathan, S. A., Bachman, P. E., & Carlson, K. (2026). qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization. The Journal of Open Source Software, 11(125), 10867. https://doi.org/10.21105/joss.10867

MLA 9

Renganathan, S. Ashwin, et al. “qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization.” The Journal of Open Source Software, vol. 11, no. 125, 2026, p. 10867. https://doi.org/10.21105/joss.10867.

Chicago

Renganathan, S. Ashwin, Peter E. Bachman, and Kade Carlson. 2026. “qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization.” The Journal of Open Source Software 11 (125): 10867. https://doi.org/10.21105/joss.10867.

Harvard

Renganathan, S. A., Bachman, P. E. and Carlson, K. (2026) ‘qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization’, The Journal of Open Source Software, 11(125), p. 10867. doi:10.21105/joss.10867.

IEEE

S. A. Renganathan, P. E. Bachman, and K. Carlson, “qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization,” The Journal of Open Source Software, vol. 11, no. 125, p. 10867, 2026, doi: 10.21105/joss.10867.

BibTeX

@article{renganathan2026qpots,
  title     = {{qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization}},
  author    = {Renganathan, S. Ashwin and Bachman, Peter E. and Carlson, Kade},
  journal   = {The Journal of Open Source Software},
  year      = {2026},
  volume    = {11},
  number    = {125},
  pages     = {10867},
  doi       = {10.21105/joss.10867},
  url       = {https://doi.org/10.21105/joss.10867}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Computational Theory and Mathematics › Advanced Multi-Objective Optimization Algorithms

  • sample efficiency
  • expensive black-box optimization
  • Python library
  • multi-objective optimization

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Renganathan, S. A., Bachman, P. E., & Carlson, K. (2026). qPOTS: A Python Package for Sample-efficient Batch Constrained Multiobjective Bayesian Optimization. The Journal of Open Source Software, 11(125), 10867. https://doi.org/10.21105/joss.10867

How many times has this paper been cited?

0 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at The Journal of Open Source Software: sign in and open the PDF from this page.

Who wrote this paper?

S. Ashwin Renganathan, Peter E. Bachman, Kade Carlson.

What is the DOI of this paper?

10.21105/joss.10867, permanently at https://doi.org/10.21105/joss.10867. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.