Implementation of attenuation maps

Closes LHCBSCIFI-90, LHCBSCIFI-89, and LHCBSCIFI-33

See merge request !101

(cherry picked from commit e674b50d)

bede3de9 Remove hardcoded attenuation tool in MCFTPhotonMonitor
8817d95e Add mirror reflectivity as paramter.
1ddbabc2 Add support for unirradiated and irradiated maps.
03bb4265 Merge branch 'master' into dev-scifi-attenuationmap
12ff16e6 Make photon map pretty
5f27ef29 Switch off wavelength generation
bbb9364e Switch on G4 attenuation maps at 50/fb
cd2dedd4 Increase number of photons for noise clusters of size 3 and 4
1fc7b4b2 Added counter for missed MCHits
cbceb8ee Rewritten PDE paramtrisation
3745b163 Fixed bug that attenutation map was returning zeros outside of boundary
4476bc35 Set mirror reflectivity to 85%
fe04356b Removed improved simulation (obsolete now)
44de2629 Changed photons/mev, relfectivity, and improved monitoring
21b506a3 Improved monitoring tools
fc4d17e6 More minor changes to the monitoring
7deb6812 Added noise deposits in neighbouring channels of a signal cluster
13b7724a Fixed potential bug
c174378d Fixed bug in adding noise deposits
2c2707cf Minor fix
7fccb2df Fixed calculation of overflow
faef4169 Fall back on old map or handle gracefully when conditions are not existing
d3e72fec Fixed wrong location of fallback attenuation map
6e27dabd Removed useless fallback
bd7cabfc Fixed wrong selection of irradiated attenuation map
82f6e550 Increased range for efficiency histogram
30488cd6 Updated values for correlated noise (minor change)
08c860fb Update photons/MeV and xtalk after testbeam tuning Luca
d935b399 Update to latest tuning from Luca
c33a861e Write main performance number in FTClusterMonitor::finalize()

- replaced `DECLARE_*_FACTORY` with `DECLARE_COMPONENT`
- made constructors public
- removed some unnecessary _disabling_ of copy constructors
- removed some trivial destructors

See gaudi/Gaudi!420

All monitors and DigitCreator compile with GaudiFunctional, DepositCreator and ClusterCreator still don't

Changed Digits to store (float)PE charge alongside the ADC thresholds. Changed the threshold values to 0,1,2,3 (indexes of thresholds, rather than some kind of ADC charge equivalent of 0, 3, 5, 9). ADCGain is now irrelevant so removed it. Changed clustering accordingly, and added PE charge weights to each threshold (for interpretation of cluster charge). Added the option in the clustering to use the digit PE charge rather than the ADC thresholds to cluster (more comparible with testbeam higher bits ADC). Changed cluster charge to be float rather than int, to accomodate for this.

BIG: implemented new noise (digit level --> deposit level). Changed cluster max charge check to match FPGA clustering


git-svn-id: svn+ssh://svn.cern.ch/reps/lhcb/Boole/trunk@197568 4525493e-7705-40b1-a816-d608a930855b