Design matrix for 2x2 factorial
In this 2x2 factorial experiment to investigate the effect of drought on tree growth, 2 different types of Populus tree were grown with 2 different amounts of water. In order to have one model coefficient per group, you need to first combine the two variables.
Cet exercice fait partie du cours
Differential Expression Analysis with limma in R
Instructions
The ExpressionSet object eset with the Populus data has been loaded in your workspace.
Combine the variables
type(the type of tree) andwater(normal vs. drought) into a single factor variable.Use
model.matrixto create a design matrix with no intercept.
Exercice interactif pratique
Essayez cet exercice en complétant cet exemple de code.
# Create single variable
group <- with(___(eset), paste(___, ___, sep = "."))
group <- factor(group)
# Create design matrix with no intercept
design <- model.matrix(~___ + ___)
colnames(design) <- levels(group)
# Count the number of samples modeled by each coefficient
colSums(design)