Post-hoc test for a zero inflated continuous data set with a tweedie
distribution?
I have a zero inflated continuous data set of aphid feeding duration on 10+
species of plant. I have fitted a glm model with a tweedie distribution and
used anova() function to show that there is significance between the plant
species. However, I would now like to perform of post-hoc test, ideally a
Tukey-Kramer type test as the data samples are of unequal size(some of my
replicates failed).
Im finding it difficult to work out id there is any appropriate
software/packages available to me in R to use on my already fitted model.
For post-hoc tests most I speak to people seem to use the tukeyHSD function
which is unusable to me as my data is unequal nor can I used aov() as I
believe my data requires glm (but I may be wrong in this). I have come
across the DTK package (short for: Dunnett-Tukey-Kramer Pairwise Multiple
Comparison Test Adjusted for Unequal Variances and Unequal Sample Sizes)
which on the face of it seem perfect, with the potentially useful DTK.test
(), but the help file is a little terse for my knowledge base so Im unsure
if it is actually possible to apply it to a fitted model with an all
important tweedie distribution.
Any help or general pointers I would be extremely grateful of
--
David Hopkins
Animal and Plant Sciences
University of Sheffield
Sheffield
UK
[[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.
