Open RiboRings opened 1 year ago
Sounds good to me at least. Should be fine when it is clear in the mia and OMA examples. Simplifying code is useful too.
That would lead subset of TreeSE which is not the optimal case. Currently, the output of runRDA includes all the samples (samples with missing values are added to vegan::rda result)
One option is to catch the na.action parameter and give more informal error message.
@RiboRings could you check the status of this one, if we can handle & close the issue. This seems straightfwd.
Ok I'll have a look
Hi!
What do you think about removing the
na.action
argument fromrunRDA
and instead providing instructions in OMA how users can filter missing values themselves?This could simplify the code of
runRDA
and prevent bugs due to differing dimensionality between tse and rda object.