experiment-design

experiment-design: Tools to create and extend experiment plans

experiment-design allows you to create high-quality experimental designs with just a few lines of code. Additionally, it allows you to extend the designs of experiments by generating new samples that cover the parameter space as much as possible...

...create and optimize orthogonal sampling designs with any distribution supported by scipy.stats...

...and easily simulate correlated variables.

And there's even more! For more details, check out the documentation and especially the section "Why choose experiment-design?".

Also, see demos to understand how the images above were created.

Install

experiment-design can be easily installed from PyPI using:

pip install experiment-design

Citing

You can cite the code using the Zenodo DOI (10.5281/zenodo.14635604). If this repository has assisted you in your research, please consider referencing one of the following works:

• Journal paper about locally extending experiment designs for adaptive sampling:

@Article{Bogoclu2021,
  title       = {Local {L}atin hypercube refinement for multi-objective design uncertainty optimization},
  author      = {Can Bogoclu and Tamara Nestorovi{\'c} and Dirk Roos},
  journal     = {Applied Soft Computing},
  year        = {2021},
  arxiv       = {2108.08890},
  doi         = {10.1016/j.asoc.2021.107807},
  pdf         = {https://www.sciencedirect.com/science/article/abs/pii/S1568494621007286},
}

• PhD thesis:

@phdthesis{Bogoclu2022,
  title       = {Local {L}atin hypercube refinement for uncertainty quantification and optimization: {A}ccelerating the surrogate-based solutions using adaptive sampling},
  author      = {Bogoclu, Can},
  school      = {Ruhr-Universit\"{a}t Bochum},
  type         = {PhD thesis},
  year        = {2022},
  doi         = {10.13154/294-9143},
  pdf         = {https://d-nb.info/1268193348/34},
}