Anne-Christine Mupepele
2010-Sep-16 14:31 UTC
[R] Nemenyi test as a post-hoc test to Kruskal Wallis
Dear all, I've discovered the possibility to do the Nemenyi-Damico-Wolfe-Dunn test in the library(coin); oneway_test() With the given example I am unfortunately not able to reproduce the test. What does trafo and contrMat mean? I have a dataframe with 176 elements in 7 classes. It may be a problem, that my model isn't balanced? In Class1 I happen to have 4 elements, while there are 90 in class2. Thank you for your help, Anne ___________________________________________________________ GRATIS: Spider-Man 1-3 sowie 300 weitere Videos!
On 2010-09-16 8:31, Anne-Christine Mupepele wrote:> Dear all, > I've discovered the possibility to do the Nemenyi-Damico-Wolfe-Dunn test in the library(coin); oneway_test() > With the given example I am unfortunately not able to reproduce the test. What does trafo and contrMat mean? > I have a dataframe with 176 elements in 7 classes. It may be a problem, that my model isn't balanced? In Class1 I happen to have 4 elements, while there are 90 in class2. > Thank you for your help, > AnneI don't quite understand your question; do you mean that you can't run the example in pkg:coin? Works fine for me. As to trafo and contrMat, try this: ?trafo ## in pkg:coin ?contrMat ## in pkg:multcomp -Peter Ehlers
Possibly Parallel Threads
- Unexpected results using the oneway_test in the coin package
- Different result with "kruskal.test" and post-hoc analysis with Nemenyi-Damico-Wolfe-Dunn test implemented in the help page for oneway_test in the coin package that uses multcomp
- Nonparametric Tukey-type multiple comparisons "Nemenyi" test
- kruskal wallis post hoc test in R
- Post-hoc test for Kruskal-Wallis