An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-mixed-models/attachments/20111112/d80e1d37/attachment.pl>
GLMM with negatively skewed data distribution
3 messages · Balázs Lestár, Reinhold Kliegl
How about scale inversion, i.e. relabeling 0:100 to 100:0? Reinhold Kliegl 2011/11/12 Bal?zs Lest?r <lestar70 at gmail.com>:
Hi All, I've just met the first negatively skewed distribution of my life! Challenging. Looks exactly like a Poisson, but with a left skew. I have discrete data points between 0 and 100. I tried to look for distribution with left skew, but I couldn't find any more than: 1. Yeah, they exist, but they are very rare. 2. Transformations like: log(max(x)+1 - x) or sqrt(max(x)+1 - x) might help. They don't in my case and also I think a GLM with the a fitting distribution would be much more appropriate. Does anyone know how to implement negatively skewed distributions in GLMMs? Bests, Balazs ? ? ? ?[[alternative HTML version deleted]]
_______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
1 day later
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-mixed-models/attachments/20111113/0fc79482/attachment.pl>