This Package focuses on multivariate robust Guassian linear regression.We provide a function Robust_Mahalanobis_regression which enables to obtain robust estimates of the parameters of Multivariate Gaussian Linear Models with the help of the Mahalanobis distance, using a Stochastic Gradient algorithm or a Fix point.This is based on the function Robust_Variance which allows to obtain robust estimation of the variance, and so, also for low rank matrices (see Godichon-Baggioni and RObin (2024) )Robust methods for estimating the parameters of multivariate Gaussian linear models. .