Use double to store energy in hits

[ManuelHu]

So we do not pretend that we have more precision in output than we actually have...

[gipert]

Uhm this needs some more discussion, we have been using floats in the past to reduce disk space usage and I'm not sure we need double precision...

[tdixon97]

I guess we need double for positions and times but not energies?

[gipert]

We need doubles only if we need to store very small or very big numbers precisely, so this is not the case for positions (and also energy...?)

[tdixon97]

Could matter for positions, eg. for computing O(um) dead layer with detectors far from the origin of the coordinate system?

[ManuelHu]

This PR was not about storage and had no changes for the output files. We already have macro options to either use double or floats in output files.

This was only about the in-memory part: It did not really make sense to have the option to write a double to disk, if we do not even have a double in the first place...

[gipert]

Sorry I misunderstood, then no problem!