Silence PMGTruthWeight tool warning on empty systematic
This is marked as a draft because I don't completely understand the implications. At least since !57599 (merged) we've been getting a warning like
PMGTruthWeightAlg.truthWeightTool WARNING Mapping for missing, setting to index 0.it's not clear if this is a problem but quite a few analyses are now ignoring it. As part of a campaign to clean up (hopefully) spurious warnings (easyjet/easyjet#52) I'm suggesting that we ignore this when the systematic is called ''. But I don't really know the implications, maybe @akong, who was responsible for the MR that last touched this, has a better idea.