One thing I would try is to supply 1/sd(y) as the weights to smoothers that can accept them; e.g., loess().
Andy > From: Matt Gibbs > > Hi, > > I have a set of observations (x,y), derived from a previous > estimation. > For each observation I also have an estimated variance s(y) > derived from > the first stage. > > The problem is that I need to smooth the data (x,y) while taking into > account the fact that the y's have been estimated at a previous stage > and thus already come with a variance. So, if I smooth the data I > somehow need to take into account *two errors*, one from the > smoothing > and the other from the already noisy data that I start off with. > > Does anyone have any idea how to do this? > > thanks, matt. > > ______________________________________________ > [EMAIL PROTECTED] mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide! > http://www.R-project.org/posting-guide.html > > ______________________________________________ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html