use Array1;
use crateOlsFit;
/// Standardized residuals `eᵢ / s`, where `s` is the residual standard error.
///
/// This is the simplest scaling — it divides by a single global estimate and,
/// unlike the studentized forms, does **not** correct for the differing
/// variances of individual residuals (via leverage). For influence work and
/// QQ plots prefer [`super::internally_studentized_residuals`] or
/// [`super::externally_studentized_residuals`].