Package: robscale 0.5.4

1. The Robustness Problem | 2. Getting Started | Installation | Quick demo | 3. The Estimator Spectrum | Decision guide | 4. The scale_robust() Dispatcher | Ensemble mode ($n < 20$) | Automatic switch to GMD ($n \geq 20$) | Explicit method selection | 5. Individual Estimators | sd_c4(x): Bias-corrected standard deviation | gmd(x): Gini mean difference | adm(x): Average deviation from the median | qn(x): Rousseeuw--Croux $Q_n$ [@rousseeuw1993] | sn(x): Rousseeuw--Croux $S_n$ | iqr_scaled(x) and mad_scaled(x) | robLoc(x): M-estimator for location | robScale(x): M-estimator for scale | 6. Confidence Intervals | 7. Performance Architecture | SIMD vectorization | $O(n)$ median selection | Stack-allocated memory arenas | Iteration convergence | TBB parallelism and ensemble design | Numerical stability and compile-time optimizations | 8. Mathematical Foundations | Logistic psi function | Decoupled estimation | Rho function for scale | $Q_n$ and $S_n$ statistics | $c_4(n)$ correction factor | Gini mean difference | Ensemble weighting formula | Key constants (full double precision) | 9. Compatibility | revss API | robustbase API | Numerical equivalence | 10. References