[R] (Cross-)validating GLM including factors
Dear list, I would like to validate various GLMs, all of them containing (among others) several factors as predictors. I thought about using cv.glm from the boot package, but necessarily got the error factor xy has new level z. Is there any function to easily cross validate LMs or GLMs containing factors that I have missed, or any other work around? - Having in mind that using cross validation on a small data set is not the best solution (I have 50-600 samples, but also three factors with up to 15 levels): Are there any other approaches? Any help is greatly appreciated! Roland [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
Re: [R] Conditional coloring of area between curves
Hi Duncan, that's exactly what I was looking for! The series have common times, so that's no problem. Thanks a lot! Roland Von: Duncan Murdoch [murdoch.dun...@gmail.com] Gesendet: Donnerstag, 6. Juni 2013 18:45 An: Roland Pape Cc: r-help@r-project.org Betreff: Re: [R] Conditional coloring of area between curves On 06/06/2013 10:41 AM, Roland Pape wrote: Dear list, I have two time series of temperatures from different sites plotted in the same diagram and would like to color the area between the curves differently, dependent on whether site 1 is cooler than site 2 (colored let's say in blue) or warmer (red). While polygone() works fine to color the enclosed area in general, I'm struggling with this conditional coloring. Any help is greatly appreciated! Suppose the series are named site1 and site2, and they have common times in a variable times. Then the following should do what you want: smaller - pmin(site1, site2) plot(x, site1, ylim = range(c(site1, site2)), type=n) polygon(c(x, rev(x)), c(smaller, rev(site1)), col=red) polygon(c(x, rev(x)), c(smaller, rev(site2)), col=blue) If the times for the two series are different it's a little harder; first you need to give them common times, and that will depend on how you decide to evaluate the values between observations. Probably linear interpolation (using approx()) is fine, but it's up to you. Duncan Murdoch __ R-help@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.
[R] Conditional coloring of area between curves
Dear list, I have two time series of temperatures from different sites plotted in the same diagram and would like to color the area between the curves differently, dependent on whether site 1 is cooler than site 2 (colored let's say in blue) or warmer (red). While polygone() works fine to color the enclosed area in general, I'm struggling with this conditional coloring. Any help is greatly appreciated! Roland [[alternative HTML version deleted]] __ R-help@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.