There is one important difference between reconstructing the real data and MC simulations for the NEMO 3. For the MC reconstruction, one need to supply list of runs to be used for efficiency degradation according to the real detector conditions (i.e. bad PMs, GG HV etc.). The way it is implemented now is that the whole MC file is splitted on the several pieces, one for each run, in proportion to the runs duration. For every piece conditions during the particular run (taken from the NEMODB) are applied. Thus every MC event is attached to a real run (variable run in the H10 tree, see later about H10).
In the end, user has a freedom to choose list of runs for the final analysis, thus it is important to ensure, that all the runs from some broad period are present in the runlist, during the MC reconstruction. If not, that would lead to the minor incorrectness of the efficiency calculations. In short, detector efficiency for the missing runs will be assumed to be equal to the average detector efficiency during the period analysed.