I've used breakaway on phyloseq object and wanted to check the model,
ba$model
but found it return
NULL
I wrote out a table and saw that my data are actually a mixture of models, mostly Neg Binomial, but some Kemp and few Poisson. Assuming breakaway acts independently on each sample, is it suspicious that there are multiple models in a single dataset? Is there an interpretation to be had here regarding the data structure?
Also, does this affect downstream betta() ? Or is error just error and missing data is just missing data at that point?
Hi Amy,
I've used breakaway on phyloseq object and wanted to check the model,
ba$modelbut found it returnNULLI wrote out a table and saw that my data are actually a mixture of models, mostly Neg Binomial, but some Kemp and few Poisson. Assuming breakaway acts independently on each sample, is it suspicious that there are multiple models in a single dataset? Is there an interpretation to be had here regarding the data structure?
Also, does this affect downstream
betta()? Or is error just error and missing data is just missing data at that point?