numeric vectors or matrices for the predictors and the response in
a linear model. Typically, but not necessarily, x will be constructed by
one of the fitting functions.
Value
The bare bones of an ols object: the coefficients, residuals, fitted values,
and some information used by summary.ols.