D-Optimal Design Calculator
Here you can create a custom D-optimal design online. Define your numeric or categorical factors, enter the available levels, choose the number of runs and select the model you want to estimate. You can also exclude forbidden factor combinations.
D-Optimal Design
A D-optimal design is a design of experiments (DoE) method that selects an efficient set of experimental runs from a larger set of possible candidates. It is useful when a standard factorial or response surface design does not fit the practical constraints of an experiment.
When to Use a D-Optimal Design
- The available factor levels do not form a standard design.
- Some combinations of factor levels are impossible, unsafe, or too expensive.
- You need to limit the number of experimental runs.
- Your experiment includes both numeric and categorical factors.
How to Create a D-Optimal Design
- Select Custom Optimal Design in the DoE calculator.
- Add your factors, select their types, and enter the available levels.
- Choose the number of runs and the model that should be estimated.
- Select D-optimal as the criterion.
- Add forbidden combinations if some candidate runs must be excluded.
- Generate the design and export the resulting test plan to Excel.
If your main goal is prediction over the candidate space, use an I-optimal design. For a broader overview of both criteria, see the custom optimal design calculator. For experiments without special constraints, you can also create a full factorial design or a fractional factorial design.