The idea of this strategy is as a replacement for !27797

After chatting with John Chapman, we realized that we could set up a
truth strategy that only modifies the keep children setting of the truth
incident. This does *not* modify the decision about the strategy (keep
or throw away the vertex); it only modifies the setting of keep
children. This can be used on top of existing truth strategies to change
their behavior, for example by keeping all the decay products of an LLP
or keeping all the decay products of a hadronic interaction, even in a
mode where they would normally be dropped. Two example configurations
are given.

At Sven's request, adding a test for the simulate cavern flag before
adding the eta/phi filter. I can't think of any situation in which we
would want to simulate the cavern but we would not want to simulate
forwward particles. After discussion with John, he agreed with the
addition.

Petr Jacka authored 5 years ago and Ruth Pottgen committed 5 years ago

Using get_E_hit function in TFCSLateralShapeParametrizationFluctChain and TFCSLateralShapeParametrizationHitChain

Moved rnd definition to L46 to select same energy bin for single/multiple sliced energy param file creation

Adjust TFCSLateralShapeParametrizationHitChain to better work with large weights and stop enforcing normalization

When running in AthenaMT, we only want to load the parametrization once.
For compatibilty with AthenaMT running the simulators themselves have to be AlgTools
though, so it makes sense to load the parametrization in a separate Service
which will then only be loaded once.

Making this change in the 21.0 branch initially, to make it easier to sweep changes to the parameterization
between 21.0 and master in the future.