Time-dependent fit model

[Mark Whitehead]

Include a fit model for performing time-dependent Dalitz-plot analysis.

Overhaul existing time-dependent model (in BaBar version of code) to fix any bugs (check SCF implementation) and apply changes to speed up and make backgrounds less B-factory specific (as has been done in the other fit models).

Add to a branch along with time pdf. Create ticket for Rafael to make both more general (i.e. add hyperbolic terms etc.) Once all completed merge into the trunk for new version.

[Thomas Latham]

(In [108]) First signal-only version of time-dependent code, with example - addresses #2

[Thomas Latham]

(In [164]) Work round compilation error on g++46 - is in resolution function code, which isn't properly working yet anyway. addresses #2 (Commits [128], [129], [145], [147], [148], [159], [160], [161], [163] also address this ticket.)

[Thomas Latham]

(In [166]) Recalculation of the interference term normalisation was not being recalculated at each fit iteration. addresses #2

[Thomas Latham]

(In [167]) Mixing phase was not being updated before the new call to calculate the interference term integral, addresses #2

[Thomas Latham]

(In [170]) Fix bug in decay time PDF that causes seg fault when not using ExpHypTrig, addresses #2