Hi everyone, I am interested in estimating this type of random effects panel: y_it = x'_it * beta + u_it + e_it u_it = rho * u_it-1 + d_it rho belongs to (-1, 1) where: u and e are independent and normally zero-mean distributed. d is also independently normally zero-mean distributed. So, I want random effects for group i to be correlated in t, following an AR(1) process. I am using the mle command, including correlation=corAR1: lme(asis~prec+pobl+gola+entr,random=~1|codi,correlation=corAR1(0.8 ,form=~temp|codi))) i = codi t = temp I am not sure whether the AR(1) process is applied to the random effects (u_it) or the error term (e_it)... Any idea? Thanks. G -- Guillermo Villa Universidad Carlos III de Madrid Business Economics Department Office 6.0.26 Madrid, 126 28903, Getafe (Madrid) SPAIN Email: guillermo.villa@uc3m.es Phone: (+34) 916249772 Mobil: (+34) 655112743 Fax: (+34) 916249607 Skype: guillermo.villa Website: www.guillermovilla.com [[alternative HTML version deleted]]
I have a question, maybe it's better to explain by example: alpha <- 0.3 beta <- 0.4 sigma <- 0.5 err <- rnorm(100) err[15] <- 5; err[25] <- -4; err[50] <- 10 x <- 1:100 y <- alpha + beta * x + sigma * err ll <- lm(y ~ x) plot(ll) Now, the graphs clearly show that 15, 25 and 50 are the indexes of the bad points. How can I retrieve this information from ll? Alberto Monteiro
Hi you can check ?influence or ?influence.measures to evaluate some regression diagnostics Regards Petr On 16 Mar 2007 at 9:56, Alberto Monteiro wrote: From: "Alberto Monteiro" <albmont at centroin.com.br> To: r-help at stat.math.ethz.ch Date sent: Fri, 16 Mar 2007 09:56:56 -0200 Subject: [R] Bad points in regression> I have a question, maybe it's better to explain by example: > > alpha <- 0.3 > beta <- 0.4 > sigma <- 0.5 > err <- rnorm(100) > err[15] <- 5; err[25] <- -4; err[50] <- 10 > x <- 1:100 > y <- alpha + beta * x + sigma * err > ll <- lm(y ~ x) > plot(ll) > > Now, the graphs clearly show that 15, 25 and 50 are the indexes > of the bad points. How can I retrieve this information from ll? > > Alberto Monteiro > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide > http://www.R-project.org/posting-guide.html and provide commented, > minimal, self-contained, reproducible code.Petr Pikal petr.pikal at precheza.cz
(I am posting the same question again, as some has replied my previous question with his own question...) Hi everyone, I am interested in estimating this type of random effects panel: y_it = x'_it * beta + u_it + e_it u_it = rho * u_it-1 + d_it rho belongs to (-1, 1) where: u and e are independent and normally zero-mean distributed. d is also independently normally zero-mean distributed. So, I want random effects for group i to be correlated in t, following an AR(1) process. I am using the mle command, including correlation=corAR1: lme(asis~prec+pobl+gola+entr,random=~1|codi,correlation=corAR1(0.8 ,form=~temp|codi))) i = codi t = temp I am not sure whether the AR(1) process is applied to the random effects (u_it) or the error term (e_it)... Any idea? Thanks. G -- Guillermo Villa Universidad Carlos III de Madrid Business Economics Department Office 6.0.26 Madrid, 126 28903, Getafe (Madrid) SPAIN Email: guillermo.villa@uc3m.es Phone: (+34) 916249772 Mobil: (+34) 655112743 Fax: (+34) 916249607 Skype: guillermo.villa Website: www.guillermovilla.com -- Guillermo Villa Universidad Carlos III de Madrid Business Economics Department Office 6.0.26 Madrid, 126 28903, Getafe (Madrid) SPAIN Email: guillermo.villa@uc3m.es Phone: (+34) 916249772 Mobil: (+34) 655112743 Fax: (+34) 916249607 Skype: guillermo.villa Website: www.guillermovilla.com [[alternative HTML version deleted]]