For Question 2.1, do we need to make our OLS functions robust to missing data in the inputs?  Or can we just assume that the dependent variable vector and the explanatory variable matrix will be complete?

Thanks,
Colin