Olá, <div><br></div><div>preciso ajustar o seguinte modelo de estimação robusta, através do método de "Mallows' para o modelo de regressão logística que segue:</div><div><div><br></div><div><br></div><div>require(robust)</div>
<div>DM.glmRob3<-glmRob(DM ~ INFLAMACAO + IDADE + SEXOM + RACACOR + HIPERT + IMC + RCQ + TRIGT, family=binomial, data=DM, weights=NULL,</div><div>        method ="mallows", model = TRUE, control = glmRob.control)</div>
<div><br></div><div>Porém, ao rodar, o que retorna é o seguinte erro:</div><div><br></div><div><div>'The covariance matrix has become singular during</div><div>the iterations of the MCD algorithm.</div><div>There are 284 observations (in the entire dataset of 400 obs.) lying on the hyperplane with equation a_1*(x_i1 - m_1) + ... + a_p*(x_ip - m_p) = 0 with (m_1,...,m_p) the mean of these observations and coefficients a_i from the vector a <- c(1, 0, 0, 0, 0, 0, 0, 0) Erro em solve.default(V)  rotina Lapack dgesv: sistema é exatamente singular'</div>
</div><div><br></div><div>Alguém já trabalhou com isso?Sabe me ajudar?</div><div>Obrigada, Natália</div>
</div>