hfsdoc

@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@                                                                                                       @
@    @@     @@ @@@@@@@@ @@@@@@@@     @@@@@@@@    @@@     @@@@@@  @@@@@@@@     @@@@@@  @@@@ @@     @@    @
@    @@     @@ @@       @@     @@    @@         @@ @@   @@    @@    @@       @@    @@  @@  @@@   @@@    @
@    @@     @@ @@       @@     @@    @@        @@   @@  @@          @@       @@        @@  @@@@ @@@@    @
@    @@@@@@@@@ @@@@@@   @@@@@@@@     @@@@@@   @@     @@  @@@@@@     @@        @@@@@@   @@  @@ @@@ @@    @
@    @@     @@ @@       @@           @@       @@@@@@@@@       @@    @@             @@  @@  @@     @@    @
@    @@     @@ @@       @@           @@       @@     @@ @@    @@    @@       @@    @@  @@  @@     @@    @
@    @@     @@ @@@@@@@@ @@           @@       @@     @@  @@@@@@     @@        @@@@@@  @@@@ @@     @@    @
@                                                                                                       @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@

Table of Contents

Tools

Background Analysis - HepFastBgAnalysis.C

Back to TOC

HepFastBgAnalysis.C is a background channel analyzer based on an analysis output file and a channel list. Since all background (BG) channels generated with the phsp generator have a unique channel number stored in the branch chan (if storage option evt is set), the tool uses this information to identify and print the channels sorted by frequence of appearance. The macro HepFastBgAnalysis.C(<opt>) expects one parameter string opt with the parameters

We can demonstrate the function by running the demo-analysis cfg/demo_jpsi.cfg containing a background source for a sufficiently large number of events with

root -l 'HepFastSim.C+(50000,"cfg/demo_jpsi.cfg")'

and afterwards run HepFastBgAnalysis.C with

root -l 'HepFastBgAnalysis.C("fana=ana_demo_jpsi.root : tree=ntp2 : fpar=parms/ftf_pbp.dat : cut=xm>2.8")'

This results in the following output:

Processing HepFastBgAnalysis.C("fana=ana_demo_jpsi.root : tree=ntp2 : fpar=parms/ftf_pbp.dat : cut=xm>2.8")...
id=1017 : (  8)  anti-p- pi+ n0 
id=1018 : (  6)  anti-n0 pi- p+ 
id=1022 : (  6)  anti-p- pi- pi+ p+ 
id=1006 : (  3)  pi- pi- pi+ pi+ 
id=1001 : (  2)  pi- pi- pi0 pi+ pi+ 
id=1002 : (  1)  pi- pi0 pi0 pi+ 
id=1003 : (  1)  pi- pi- pi0 pi0 pi+ pi+ 
id=1011 : (  1)  anti-p- pi0 p+ 
id=1012 : (  1)  anti-n0 pi- pi0 p+ 
id=1013 : (  1)  anti-p- pi0 pi+ n0 
id=1016 : (  1)  pi- pi0 pi0 pi0 pi0 pi+ 
id=1020 : (  1)  pi- pi- pi- pi0 pi+ pi+ pi+ 
id=1023 : (  1)  anti-p- pi0 pi0 p+ 
id=1045 : (  1)  pi- pi+ eta

We e.g. see that the channel anti-p- pi+ n0 is the most dominant background channel in the selected mass region m > 2.8.

PID Function Cache - HepPidHistGen.C

Back to TOC

The macro HepPidHistGen.C offers the functionallity to generate a PID histogram map (see parameter pidmap in PID detector setup from input functions in ROOT’s TFormula format. These functions should depend on mass m and momentum p and might depend on angles theta (tht) and phi. The code will generate a set of TH3F, one for each particle type (e, mu, pi, K, p), with the axis x=p, y=theta, and z=phi, with the bin contents and bin-errors being the expectation values and resolutions, respectively, of the PID quantity for the corresponding phase-space location (p, theta, phi). The macro HepFastBgAnalysis.C(<opt>) expects one parameter string opt with the parameters

The functions fexp, fres, and fthr can make use of an arbitrary set of additional parameters, which need to be defined in opt. Depending on the complexities of the functions, caching by corresponding histograms could speed up the simulation a little, but at the expense of granularity (binning). The generated histograms have the names <name>_<particle> for all particles in the list part, where <particle> is the particle name without the charge symbol, i.e. e, mu, pi, K, and p. The macro allows to add histograms to an existing files, so that in histogram maps for multiple detectors can directly by stored in one file.

Let us for example create histograms for a Time-of-flight detector with the corresponding PID quantity being the velocity beta = v/c = p/E. The macro call could look like this:

root -l -b -q 'HepPidHistGen.C("fexp=[p]/sqrt([p]^2+[m]^2) : fres=[p]^2*[dt]*30*sin([tht])/(([p]^2+[m]^2)*[dist]) : dist=100 : dt=0.1 : name=tofb : htht=100,20,140")'

which generates the histograms tof_e, tof_mu, tof_pi, tof_K, and tof_p with axis parameters x=(100,0,6), y=(100,20,140), and z=(1,-180,180) inside the file histpid.root for a barrel shaped ToF detector. The additional parameters dist (radial distance of the ToF system to the interaction point in cm) and dt (time resolution of the detector in ns) are needed for the resolution function fres. The factor 30 is the speed of light in units cm/ns. The generated histogram maps can be used in the simulation configuration with

PID ;; name=tof_barrel : pidmap=histpid.root, tofb

like pointed out in this section.

PID Pdf Generator - HepPidPdfGen.C

Back to TOC

In case the probability density (PDF) for a certain PID quantity is not Gaussian distributed (see figure in the PID section for examples), there is the possibility to provide the PDF in form of a TH3F histogram, where the (x,y) bin corresponds to the phase-space location (p, theta), and the distribution along the z-coordinate represents the normalized PDF for the quantity. The macro HepPidPdfGen.C helps to generate these kind of histograms based on an input TTree inside a ROOT file.

In addition it offers an interactive display of histograms already generated (just now or read from the input ROOT file) to inspect the PDF for the various particle types. The macro HepPidPdfGen.C(<opt>) expects one parameter string opt with the : separated parameters

For an ROOT input file named pid10M.root containing the quantity muo (e.g. for the penetation depth of the particle in the iron yoke of a muon detector covering polar angles 20° < theta < 140°) and the true PDG code trpdg, a call of the macro could look like:

root -l -b -q 'HepPidPdfGen.C("fin=pid10M.root : var=muo : hvar=100,0,100 : htht=20,20,140")'

This projects the 3D-distributions (p, theta, var) in TH3F with dimensions p,theta,var = (100,0,8) x (20,20,140) x (100,0,100) and stores/updates the histograms muo_e, muo_mu, … , muo_p in the file hpidpdf.root. If we would just like to inspect distributions of a PID variable in histograms with prefix muo being already stored in a file named hpidpdf.root, the command

root -l 'HepPidPdfGen.C("fin=hpidpdf.root : var=muoiron : load : logy : logz")'

can be used. Two windows pop up, where one (named PDF) shows the 2D-projections (p, var) for all five particle types (e, mu, pi, K, p), and in the bottom right pad of this window the 1D-projections of the PDFs for all five particle types are displayed superimposed. The corresponding phase space location (p, theta) is picked in the (p, theta) plane shown in the second window (named PHSP). When moving the mouse pointer, the 1D-projection is update. By clicking in the PHSP window, the location is fixed until clicking again in the plane.

The generated histogram maps can be used in the simulation configuration with

PID ;; name=muon_det : pidpdf=hpidpdf.root, muoiron

like pointed out in this section.

Histogram Drawing - HepDrawHistos.C

Back to TOC

The macro HepDrawHistos.C offers an easy way to plot a bunch of histograms from a ROOT TTree. In particular it simply re-draws live histograms from a simulation job based on the output file and the configuration file containing the HIST statements. The macro expects up to three input parameters HepDrawHistos(cfg, file, opt)

For example, if we would like to re-draw the plots from the simulation corresponding to cfg/demo_mini.cfg, we just need to execute

root -l 'HepDrawHistos.C("cfg/demo_mini.cfg")'  // -> auto-generates file name "ana_demo_mini.root"

displaying the canvas with the two plots as shown in the introduction.

Proceed to the next section: Demos