My Home
To run in the debugger, in a new shell do:
- setenv CMTCONFIG $CMTDEB
- setenvDaVinci v19r13
- cd to directory
- cmt config then source setup, stay in directory
- cmt make
- ls ..
- ls ../slc4_amd64_gcc34_dbg
- ls ../$CMTCONFIG check they're the same
lpr -P 13-3-COR filename
maybe one can do combineparticles to do partial reconstructions
MCDecayFinder:
- Note: "->" means "decay to" and "=>" means "decay to .... with some resonnances in the middle". So if you want to catch B0 to D_s- K+ with any D*_s in beetween you simply write "B0 => D_s- K+ ...". NB: The "..." is here to catch any number of particles of any kind so you get the products of the D*_s-.
LoKi ADAMASS is a mass difference to the reference value, not an absolute mass cut
ROOT colours at
http://root.cern.ch/root/html/TAttFill.html
category0 =signal
category20 =
FullyRecoPhysBkgrd: misreconstructions or correct decays proceeding via an intermediate resonance
Always put PID tool in decaytreetuple
dll = deltalog likelihood-variables leaves are combinedlikelihood of mas hypothesis. deltalog kdll= log k - log pi
deltalog of anything is the difference of the log from that of a pion.
PIDK is the decaytreetuple syntax for kdll
To find out what names have been given to particles in a decay involving two or more of the same particle look at the first decay in the log file.
m(Kpi) = invariant (4-modulus) of sum of 4-vector momenta pf K and pi.
To write to CASTOR:
- rfio://castor/cern.ch/user/h/hgordo, or from within file FILE = '/castor/cern.ch/user/j/joel/tar/myfile.tar'
- reading from CASTOR: "rfio:/castor/cern.ch/user/j/joel/tar/myfile.tar /localdisk/joel/myfile.root", or within Gaudi: FILE = '/castor/cern.ch/user/j/joel/tar/myfile.tar'
- list directory nsls -l /castor/cern.ch/user/j/joel
- remove file nsrm /castor/...
Particles stored in $PARAMFILESROOT/data/ParticleTable.txt
Execute
DaVinci options file $DAVINCIROOT/$CMTCONFIG/DaVinci.exe file.opts
DecayTreeTuple tools in $DECAYTREETUPLEROOT/src
grep string filenameorpath | wc -l //displays number of occurrences of string (ie line-count; wc)
EXPECT roughly 81% "efficiency" for D to KPI- though this result is not easy to find and may relate to optimisation for D*.
root [4]
B2D2KKPiTuple->Draw("Dminus_MM:Dminus_BKGCAT", "", "lego")
(Long64_t)365
root [5]
B2D2KKPiTuple->Draw("Dminus_BKGCAT:Dminus_MM", "", "lego")
(Long64_t)365
Table 3 lists the selection criteria for the HLT D ∗ stream. The two oppositely charged tracks (h− h+ )
used in reconstructing the D 0 candidate must separately pass lower limits on total momentum (|p|),
pt , and IP significance (IPs) with respect to every reconstructed PV. Their combined momentum must
also have pt (D0 ) > 1.25
GeV. A common origin vertex for the two tracks (D 0 decay vertex) is cal-
culated and required to have a χ2 /Ndof < 16 and a separation significance > 6.0 with respect to
that PV to which the D0 (h− h+ ) candidate has the least IPs. A large window of D 0 (h− h+ ) invariant
masses, −700
MeV < (mh− h+ − mD0 ,PDG ) < 50
MeV, is allowed to accommodate several D 0 decay
modes. The assumed pion mass for h± leads to an underestimation of the D 0 mass and introduces a
significant low mass tail for the D 0 → K π ± and D0 → K − K + decay modes, as shown in Figure 1.
D0 candidates are combined with a candidate slow pion track to form D ∗+ → πs D0 candidates. The
+
πs track must have minimum IPs > 1 with respect to all reconstructed PVs. Like the D 0 (h− h+ ) can-
didate, the combined momentum of the D ∗+ (πs D0 ) must satisfy pt (D∗+ ) > 1.25
GeV, and its decay
+
vertex must have a χ /Ndof < 16. The size of the D ∗+ mass window is similar to the D 0 mass window.
2
Since the difference between the masses of the reconstructed D 0 and D∗+ candidates depends only
weakly on the D0 mass, and hence only weakly on the mass hypothesis assigned to the h− h+ products
of the D0 , the mass difference ∆m ≡ mπ+ D0 − mh− h+ remains a strong signature of D ∗+ → πs D0 de-
+
cays, even when the final state mass hypotheses are mis-assigned. The reconstructed mass difference
must fall within the window |∆m − ∆mPDG | < 10
MeV. If one or more D ∗+ candidates that satisfy
these criteria can be found in an event, that event passes the HLT D ∗ trigger stream and is saved for
physics analysis. Table 4 lists the number of events from each Monte Carlo sample that pass the L0,
L1, and HLT D∗ trigger sequence. It also shows the number of RS D 0 decays in the triggered events.
The same selection gives an estimated yield of 1.3 × 10^6
D∗+ → π + D0 (K − K + ), and 0.5 × 106 D∗+ → π + D0 (π − π + ) per 2 fb^−1 , where the D∗+ originates from
a b-hadron decay. The estimated background to signal ratio for these two modes are 0.10 < B/S < 0.35
and 0.16 < B/S < 0.78, respectively, at 90% CL. We have presented a preliminary sensitivity study of
the mixing parameters x 2 and y , with estimated statistical uncertainties 6.4 × 10−5 and 8.7 × 10−4
respectively in 10 fb−1 of data. Also presented is a study of the mixing parameters yCP and CP vio-
lation parameter AΓ , with estimated statistical uncertainties 4.9 × 10−4 and 4.8 × 10−4 respectively. If
the strong phase difference, δ, between the CF and DCS Kπ decays is known, and if CP violation is
assumed to be negligible, then the two sets of measurements can be combined. LHCb will measure
δ with a precision of ∼ 5◦ in 10 fb−1 [24]. Present results suggest δ is small [25]. The likelihood func-
tions from both mixing studies have therefore been combined under the assumption yCP = y . The
resulting contours are shown in Figure 23. In this combined analysis the precisions on the mixing pa-
rameters improve to σx 2 = 3.1 × 10−5 and σy = 3.8 × 10−4 . The studies presented show that LHCb
has the statistical power to measure charm mixing and CP violation with unprecedented precision.
The ultimate sensitivity will depend on how well the systematic uncertainties can be controlled.