The DOE Toolkit includes the doeDesirability function, which allows you to explore the desirability of your single/multiple response model for alternative settings of the independent variables in the model.
The doeDesirability function provides slider controls that allow you to adjust the values of the independent variables in your model. Manipulating the sliders causes the desirability function for each response variable to be recomputed and its associated ListPlots to be redrawn. This allows you to interactively explore the desirability of the multiple response model for a given set of factor settings.
Let’s assume that you have used one of the DOE Toolkit functions to create an experimental design and that you have run your experiment and recorded the results. In order to be able to use the doeDesirability function to analyze the desirability of the factor settings, you must first complete the following steps:
Define vectors for the observed effects in your model. These vectors are the observed values for the effect variables that were obtained by running your experiment.
Define vectors (derived from your design) for the factors in your model.
Create a fitted model for each of the effects in your model.
Define a list of the fitted models created in Step 3.
Define the effect-based desirability functions to be used for your model.
Set lower, target, upper values for the effects in your model.
Define factor-based desirability functions.
Define a list of the factor-based desirability functions created in Step 7.
The Doe toolkit user guide 2.0.nb notebook provides an example in which these steps are followed in order to use the doeDesirability function to analyze a multiple response model based on the example described in Derringer and Suich[1].
The screen capture below shows the doeDesirability function in use for the D&S model.
[1] George Derringer & Ronald Suich (1980) Simultaneous Optimization of Several Response Variables, Journal of Quality Technology, 12:4, 214-219, DOI: 10.1080/00224065.1980.11980968
DOE Toolkt for Mathematica © 2020 Harper Corditt Software. All rights reserved.
Mathematica® is a registered trademark of Wolfram Research.