search

dwarton / mvabund

mvabund updates
10 stars 14 forks source link

Obtain univariate p-values for all pairwise comparisons of factor levels in anova.manyglm #72

Closed kdpaepe closed 2 years ago

kdpaepe commented 4 years ago

Dear,

I was wondering if there is a way to combine the pairwise comparison option with the p.uni="adjusted" option in anova.manyglm to obtain adjusted p-values for every response variable for every possible pairwise combination of factor levels.

While p.uni="adjusted" or pairwise comparisons work fine when specified separately, I get the following error message when I use both arguments together: Error in dimnames(x) <- dn : length of 'dimnames' [2] not equal to array extent

Many thanks in advance for your time and consideration.

dwarton commented 4 years ago

I can't replicate your issue without reproducible code but I think it is unrelated to your request - pairwise.comp and p.uni="adjusted" can both be supplied and will run. However at the moment they are not both done at the same time - pairwise comparisons is done on the global test only