Class for the analysis of data obtained from a design of experiments approach. It does aim to re-invent the relevant, individual methods already implemented in SciPy, statsmodels, scikit-learn etc., but rather to provide an easy to use module that does some data preprocessing and is a collection of the most useful methods from the external packages.
- Scales the parameter ranges to the interval of [-1, 1]. Thus, the fit parameters of the regression model can be used directly to compare the effects of the various experimental parameters.
- Does the regression with linear, two-factor and three-factor interaction as well as quadratic models.
- Performs an analysis of variance (ANOVA) on the data.
- Also contains some useful methods for generation of plots for model diagnostics.
Download and run the following command from the repository folder works:
pip install -e .