Print

Print


IslBrecoilUser rf-060602
contains a class, recoilDSys, which computes the weights to be applied to
modify the B->Dlnu and D->X branching fractions. The compilation
and linking should be ok now, although you have to make sure that you have
loaded libNtuples.so in your job (Alessio, also the ntuplesLinkDef.h had
changed did you get it?).

Here is a description of what it does and how to use it. It has two
constructors and two weight functions, one for each kind of systematics
The assumed values for the DECAY.dec and PDG branching fractions are in
http://babar.roma1.infn.it/~faccini/Dsys_plots.html

a) B->Dlnu BF

The constructor to be called is the one with only one argument, the random
seed:
recoilDSys::recoilDSys( int seed)
if the seed is 0, the weights are just the ratio of the PDG over the
DECAY.dec. Otherwise random numbers are thrown (once for the whole just,
pay attention not to istantiate the same class for each event) with the
supplied seed and the PDG numbers are varied according to their resolution

the weight function
float  recoilDSys::weight(int bmode)
is to be called each event and requires the bmode, i.e.:

0     not vcb
+/- 1 (Dlnu)
+/- 2 (D*lnu)
+/- 4 (D1*lnu)
+/- 5 (D2*lnu)
otherwise (D**lnu non-res)
+/- stands for B0(+) or Bch (-)

b) D->X

the constructor requires two arguments
recoilDSys::recoilDSys(const char *name, int seed)

the name of the file that contains the list of BF one wants to consider in
the format
# name   #Pi    #K   #K0   #pi0   Br(decay.dec)    Br(PDG)  sigmaBr(PDG)
For istance the D0->K+p- looks like
Kpi       1     1     0     0      0.0387          0.0383    0.0009

seed needs to be 0 if you just want to rescale to PDG

The weight function, that assumes that only one D in the recoil
float recoilDSys::weight(int npi,int nk, int nk0, int npi0, int nlep,int & mode)
requires as inputs the number of pions, kaons etc etc in the decay and
returns the weight and the "mode" of the D decay, in case one wants to use
this classification.


	hope this helps, please report problems
	ciao
	ric