How should I set the optimization range?
Setting the optimization range to a subset of the frames allows MoMA to perform optimization faster. Hence it is desirable to use the largest optimization range at which optimization is still very fast (typically 150-300 frames). In datasets where a few consecutive frames are particularly problematic, it is useful to transiently set a very short optimization range to fix these frames (since optimization might be run many times).
I got an error message "ILP unfeasible. Please re-optimize..." What should I do?
This error happened because the constraint that you entered manually by setting a cell or an assignment can't be handled by MoMA. Before restarting the curation from scratch, it's worth resetting the corresponding frame and/or transition: this should bring you back to the state you were at before setting the faulty constraint.
Does the segmentation result depend on the optimization range that I set?
In practice, one wants to use the largest optimization range at which optimization is still very fast (typically 150-300 frames). Of course, the optimal solution found by MoMA can be impacted by missing additional information contained in future frames (close to the end of the range). The only documented case where this is problematic is the analysis of very difficult datasets where the user relies on a very short optimization range to produce a correct solution; when reloading the curation file, the optimal solution is likely to contain several mistakes since the results were checked with optimization run a local subsets of the data...