Tuomas Borman

Alright, now the tests should be ok. I believe the stochastic nature comes from multithreading and complex tree (loops); distances between tips can be calculated multiple ways which slightly differs....

Thanks! Certain ordination methods that we use come from scater package. They use calculate* and run*, for instance, calculatePCA and runMDS. That is why we have to think about if...

@ake123 Can you check these? E: at least, some. There might be more things coming.

4. https://microbiome.github.io/OMA/docs/devel/pages/98_exercises.html#assay-transformation The extra part of this exercise should say as a hint that phiLR package should be used.

5. It seems that MARGIN can be only "samples" or "features" in transformAssay(). However, in R 1 and 2 are the most common values. We should switch to 1 and...

6. mia's documentation should be fixed. `runRDA()` has hidden parameter permutations. However, the documentation does not say clearly how to pass the values to this parameter. We could improve documentation...

7. diversity functions fail when there are values that certain indices do not support (relative or negative). Instead, this should lead to a warning that tells that these indices where...

It should be, can you check why? https://microbiome.github.io/mia/reference/runCCA.html Do you have latest version?

To summarize: 1. Leave splitByRanks as alias. (just empty function calling agglomerateByRanks) 2. In SE method, add TreeSE conversion if the class is SE. (Give warning about the conversion) 3....

Yes, the purpose is not to copy those resources but rather create a platform from where to find relevant resources in microbiome data science. The focus is still on downstream...