Experimental design based on GLM model

Hello all,
I 'm very new to R. I was wondering if there was any way to predict an optimal experimental design from a table, just by running a GLM on it.
For instance, I would run a function on a 1000 observations table, specifying in input the variables of my model (the GLM), and the number of experiment I want to perform. And according to which variables are the most significant and the number of modalities of those variables, I would get a table of experiments with all the variables specified and the number of experiments specified in the function...

I hope I've maid myself clear enough ^^

I thank you very much for your response :slight_smile:

I must precise that my model contain 8 variables with around 10 levels for each variable. Only two of these 8 variables are correlated.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.