Design of Experiments (DoE) in Python?

Question

Is there any module or package available in Python for using Design of Experiments (DoE) methods?

Sitzmark

Answer 1

Have you tried pyDOE? It is great for constructing factorial designs, response surface designs, and more. You can find it on PyPi.

Cimabue

Answer 2

Try this new code base which takes a simple CSV file as input variable's range and generates the desired DOE in another CSV file. It is not a full package but collection of functions to use.

Design-of-Experiment-Python

Giuliana

Answer 3

I would suggest to use OpenTURNS's DoEs. The provided DoE are:

• stratified DoE (axial, factorial, composite, box)
• random (bootstrap, LHS, MonteCarlo, importance sampling)
• deterministic (fixed, Gauss product, tensor product, Smolyak)
• cross validation (K-Fold, leave one out)
• low discrepancy (Faure, Halton, reverse Halton, Haselgrove and, of course, Sobol')
• optimized LHS (from Monte-Carlo or simulated annealing, with different criteria)

Furthermore, there are different quadrature methods, which are another set of features for design of experiments.

It's easy to install using pypi or Conda.

Tryptophan