RICH 4D Reco: Add new algorithm to add Tk origin 4D vertex using MC to existing reco segments

Adds a new algorithm for RICH 4D reco studies, that uses MC truth to add 4D origin vertex information to existing track segments.

With this option, its possible to run the 4D RICH recontruction using fully reconstructed tracks, as well as the already existing option to cheat the tracks using RICH XSIM information.

See the example options file Rich/RichFutureRecSys/examples/RichFuture.py for examples on how to enable this option. Basically, if you set

useMCTracks = False
tkSegAddTimeFromMC = False

Then the regular reco tracks are used, but the new algorthm is also run that adds the 4D origin information.

Then, if you also enable

useMCHits = True

to get the time into the RICH hits, and finally use

# Enable 4D reco in each RICH
enable4D = (True, True)
# Average expected time for signal
avSignalTime = (13.03, 52.94)
# Pixel time window size in each RICH (in ns)
#pixTimeWindow = (99999, 99999)
pixTimeWindow = (1.0, 1.0)
# Photon reco time window size in each RICH (ns)
#photTimeWindow = (99999, 99999)
photTimeWindow = (0.25, 0.25)

the various timing options can be experimented with.

Finally, just to prove it works running the regular 3D reco I see

RichPIDMonLong2to100                       INFO Number of counters : 8
 |    Counter                                      |     #     |    sum     | mean/eff^* | rms/err^*  |     min     |     max     |
 |*"Event ID rate"                                 |      1000 |        992 |( 99.20000 +- 0.2817091)% |
 |*"K->K,Pr,D ID Efficiency"                       |      2588 |       2526 |( 97.60433 +- 0.3005842)% |
 |*"Pi->El,Mu,Pi ID Efficiency"                    |     35775 |      35221 |( 98.45143 +- 0.06528092)% |
 |*"Pr DLL(Pr)>0 ID Efficiency"                    |      1034 |        992 |( 95.93810 +- 0.6139029)% |
 |*"Track ID rate"                                 |     68175 |      64227 |( 94.20902 +- 0.08945604)% |
 |*"Used RICH1 and RICH2"                          |     64227 |      20092 |( 31.28279 +- 0.1829476)% |
 |*"Used RICH1 only"                               |     64227 |      30944 |( 48.17911 +- 0.1971619)% |
 |*"Used RICH2 only"                               |     64227 |      13191 |( 20.53809 +- 0.1594046)% |

but with the above to add the origin time to the reco track segments and rich hits, and to enable the 4D reco I see

RichPIDMonLong2to100                       INFO Number of counters : 8
 |    Counter                                      |     #     |    sum     | mean/eff^* | rms/err^*  |     min     |     max     |
 |*"Event ID rate"                                 |      1000 |        992 |( 99.20000 +- 0.2817091)% |
 |*"K->K,Pr,D ID Efficiency"                       |      1682 |       1661 |( 98.75149 +- 0.2707419)% |
 |*"Pi->El,Mu,Pi ID Efficiency"                    |     28372 |      28170 |( 99.28803 +- 0.04991536)% |
 |*"Pr DLL(Pr)>0 ID Efficiency"                    |       564 |        550 |( 97.51773 +- 0.6551288)% |
 |*"Track ID rate"                                 |     68175 |      64227 |( 94.20902 +- 0.08945604)% |
 |*"Used RICH1 and RICH2"                          |     64227 |      20092 |( 31.28279 +- 0.1829476)% |
 |*"Used RICH1 only"                               |     64227 |      30944 |( 48.17911 +- 0.1971619)% |
 |*"Used RICH2 only"                               |     64227 |      13191 |( 20.53809 +- 0.1594046)% |

FYI @fkeizer @lmalenta @seaso

Needs LHCb!4281 (merged) for an MC association bug fix