By default, regsubsets() only reports results up to the best eight-variable model. But the nvmax option can be used in order to return as many variables as are desired. Here we fit up to a 19-variable model.

regfit.full <- regsubsets(Salary ~ ., data = Hitters, nvmax = 19)
reg.summary <- summary(regfit.full)

Try using the regsubsets() function in the same fashion with the Boston dataset with the nvmax option set to 13 and store the output in regfit.full:

Assume that:

• The MASS and leaps libraries have been loaded
• The Boston dataset has been loaded and attached