Get the design matrix from a linear mixed model

Question

Armand Mensen 2016-11-4

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/310634-get-the-design-matrix-from-a-linear-mixed-model

回答： Xue Zhang 2021-5-7

When running a linear mixed model, using the fitlme function...

formula = 'dependent ~ predictor1 * predictor2';
model_object = fitlme(dataset, formula);

... you can extract the design matrix after you have run the model using the following code:

% get fixed effect design
design_fixed = designMatrix(model_object, 'fixed');

However, I'd rather not have to run the whole model every time I want to just get the design matrix. So I'm looking for a function that will output the design matrix given the dataset table and the formula used in the model. For example:

design_matrix = designFromFormula(formula, dataset, dummyvarcoding);

Precisely such a function already exists buried deep within the 'LinearMixedModels.m' at line 1392; but this function is private so I cannot access it. Nor can I re-engineer exactly what it is doing.

Does someone know how I could use this private function, or have another way to 'extract' the design matrix for a linear mixed model (just the fixed effects to be specific), without having to actually estimate the model?

Thank you!