dimitris kapetanakis
2008-Sep-30 17:53 UTC
[R] Quantile Regression for Longitudinal Data. Warning message: In rq.fit.sfn
Hi, I am trying to estimate a quantile regression using panel data. I am trying to use the model that is described in Dr. Koenker's article. So I use the code the that is posted in the following link: http://www.econ.uiuc.edu/~roger/research/panel/rq.fit.panel.R While this code run perfectly, it does not work for my data providing a warning message: In rq.fit.sfn(D, y, rhs = a) : tiny diagonals replaced with Inf when calling blkfct So I am wondering if I am doing something wrong or if it is data's problem (which runs perfectly at least for fixed and random effects as well as for quantile regression adding dummies of states and years just like fixed effects). Any help would be highly appreciate Thanks Dimitris ###################################### The code that I use before using the link is: data$lex2<-data$lex^2 data$lex3<-data$lex^3 data$constant<-rep(1,times=length(data$lex)) X<-cbind(data$lex, data$lex2, data$lex3, data$constant) y<-c(data$ban) s<-c(data$state) -- View this message in context: http://www.nabble.com/Quantile-Regression-for-Longitudinal-Data.-Warning-message%3A-In-rq.fit.sfn-tp19747218p19747218.html Sent from the R help mailing list archive at Nabble.com.
roger koenker
2008-Sep-30 22:53 UTC
[R] Quantile Regression for Longitudinal Data. Warning message: In rq.fit.sfn
This is a little esoteric for R-help. As the posting guide says, you should write the package maintainer with this sort of question. Without the data it is difficult to judge what is happening, a couple of possibilities are: o all is well and warning just conveys an exaggerated sense of skepticism about one or more of the Cholesky steps, o or, if the coefs really look like nonsense, you have a near singularity in the design, if so you can try to look in more detail at the dense form of design and try to see what the problem is, provided of course that it is small enough to look at with the usual tools, svd, etc. url: www.econ.uiuc.edu/~roger Roger Koenker email rkoenker at uiuc.edu Department of Economics vox: 217-333-4558 University of Illinois fax: 217-244-6678 Champaign, IL 61820 On Sep 30, 2008, at 12:53 PM, dimitris kapetanakis wrote:> > Hi, > > I am trying to estimate a quantile regression using panel data. I am > trying > to use the model that is described in Dr. Koenker's article. So I > use the > code the that is posted in the following link: > > http://www.econ.uiuc.edu/~roger/research/panel/rq.fit.panel.R > > While this code run perfectly, it does not work for my data > providing a > warning message: > > In rq.fit.sfn(D, y, rhs = a) : tiny diagonals replaced with Inf when > calling > blkfct > > So I am wondering if I am doing something wrong or if it is data's > problem > (which runs perfectly at least for fixed and random effects as well > as for > quantile regression adding dummies of states and years just like fixed > effects). > > Any help would be highly appreciate > > Thanks > > Dimitris > > ###################################### > The code that I use before using the link is: > > data$lex2<-data$lex^2 > data$lex3<-data$lex^3 > data$constant<-rep(1,times=length(data$lex)) > > X<-cbind(data$lex, data$lex2, data$lex3, data$constant) > > y<-c(data$ban) > > s<-c(data$state) > > > > -- > View this message in context: http://www.nabble.com/Quantile-Regression-for-Longitudinal-Data.-Warning-message%3A-In-rq.fit.sfn-tp19747218p19747218.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > R-help at r-project.org 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.