This is a documentation issue. In principle, could also make it possible for the user to select a filtered (or even transformed) instance of the data as input here, but in most cases, the rawest of counts is what they should be using, so it should be a relatively strong default in the way it is presented.
This is a documentation issue. In principle, could also make it possible for the user to select a filtered (or even transformed) instance of the data as input here, but in most cases, the rawest of counts is what they should be using, so it should be a relatively strong default in the way it is presented.