Simplify access to ensemble members and statistics #113
Description
The stats submodule contains several functions for computing specific statistics on the ensemble. However, it is still cumbersome to compute "custom" statistics. To that end, it might be useful to provide a function that takes an observation sequence as input and returns, e.g., only the ensemble of a specific phase as a data frame. Then the user can simply compute statistics "on their own" and need not filter the data frame columns themselves. Would that be a useful addition to an existing workflow?
I raise this issue as part of my review in openjournals/joss-reviews#9439