Fix Coverity leaks

[u/mrichter/AliRoot.git] / PWG3 / READMEmuon

===================================================
  Code to include the muon information in the 
  creation of the standard AOD from the ESD
  Roberta ARNALDI, Enrico SCOMPARIN Torino
---------------------------------------------------------------

The following filter is used in the official analysis train, in order to copy the muon information from the ESD to the standard AOD.

1) AliAnalysisTaskESDMuonFilter.h  --> analysis task to copy the muon information from the ESD to the standard AOD
2) AliAnalysisTaskESDMuonFilter.cxx --> analysis task to copy the muon information from the ESD to the standard AOD

In order to test the analysis train from the ESD to the Standard AOD production, the following macro is provided:

  AnalysisTrainFromESDToAOD.C

Input file: ESD
Output files: standard AOD (+AOD tags)

Two wagons are attached to the analysis train: 
   1) $ALICE_ROOT/ANALYSIS/macros/AddTaskESDFilter.C 
   2) $ALICE_ROOT/PWG3/muon/AddTaskTagCreation.C 

Several flags can be activated: 
   1) iESDfilter: to activate, using AddTaskESDFilter.C, the copy of the ESD information into the AOD.
      In AddTaskESDFilter.C there is the possibility to apply cuts on the tracks and muon tracks in order
      to reject them before filling the AOD. 
   2) iAODAddMCBranch: to activate the inclusion of the MC branch (containing Kinematics info) into a branch of the AOD 
   3) iAODTagCreation: to activate, using AddTaskTagCreation.C, the AOD tag creation
      
Runninng options tested: 
   1) GRID (with/without AliEn plugin)
   2) LOCAL 

If the AliEn plugin is required, the macro 
   CreateAlienHandler_FromESDToAOD.C
can be used, in order to configure the jdl.

AnalysisTrainFromESDToAOD.C is an updated version of the macro AnalysisTrainMuonLocal.C 
(input file: ESD, output files standard AOD and the AOD tag file)
The main differences are related to implementations to be compliant with the analysis framework,
possibility to run on the grid with/without the Alien Plugin and possibility of adding the MC truth in a branch of the AOD.

The macro ReadAOD_MCBranch.C is an example on how to access MC information stored in the AOD.
   
===================================================
  Creation of the MUON-AOD from the standard AOD
  Roberta ARNALDI, Enrico SCOMPARIN Torino
---------------------------------------------------------------

Code to produce a muon-AOD, i.e. a replica of the standard AOD containing only events where 
at least one muon in the dimuon spectrometer is present

1) AnalysisTrainFromStandardToMuonAODLocal.C --> macro to produce the MUON-AOD (it runs locally)
2) AliAnalysisTaskFromStandardToMuonAOD.h  --> analysis task to replicate the AOD content for interesting events
3) AliAnalysisTaskFromStandardToMuonAOD.cxx --> analysis task to replicate the AOD content for interesting events

- The input files are the standard AOD and the AOD tag file.
  The AOD tag file can be created in the previous step from ESD to Standard AOD, using AnalysisTrainMuonLocal.C, or it can be
  produced on the fly with this macro.
- Events containing at least one muon in the muon spectrometer can be selected using the AOD tag files.
- The contents of the MUON-AOD can be defined by the user in the AnalysisTrainFromStandardToMuonAODLocal.C
  macro with some settings as
  SetNeedsTracksBranchReplication(), SetNeedsVerticesBranchReplication() 
  (defined in STEER/AliAODHandler.h)...

===================================================
  Calculation of the LUT 
  Bogdan VULPESCU, Clermont
---------------------------------------------------------------
Analysis task for the calculation of the Look-up-Tables used by the MUON
trigger algorithm

Purpose: extract from ESD files track information (ESD track parameters and
matched local trigger decisions) and store in TNtuple; the ntuple is used
for building 2D histograms (transverse momentum / local deviation) which
afterwards enter calculation/optimization procedures for the selection of
LUT cuts for individual local boards and several transverse momenta.

This version is prepared for the use on the grid, see the JDL file.

Files:

AliAnalysisTaskLUT.cxx   - source files of the analysis task
AliAnalysisTaskLUT.h

rootlutnt.sh             - executable (script)

runCreateLUTNT.C         - macro for setting the libraries environment and
                           execution of the main macro

analysisTaskLUTNT.C      - main macro for the analysis

createLUTNT.jdl          - JDL file

===================================================
 Analysis task for the calculation of the trigger chamber efficiency.
 Diego STOCCO, Torino
----------------------------------------------------------------

Purpose: produce histograms for the trigger chamber efficiency determination 
from MUON track info. The task works either with input ESD or AOD (default is ESD).

This version works either locally or on grid.

Files:
AliAnalysisTaskTrigChEff.cxx   -source files of the analysis task
AliAnalysisTaskTrigChEff.h

AnalysisTrigChEff.C            - main macro for analysis

Test analysis:
- Preliminary: copy the following packages in the current working directory
  - STEERBase.par
  - ESD.par
  - AOD.par
  - ANALYSIS.par
  - ANALYSISalice.par
  - PWG3muon.par
If the packages are not present, create them: 
cd $ALICE_ROOT
make STEERBase.par
make ESD.par 
...

  - Local test:
  root
  .x AnalysisTrigChEff.C(kMlocal)

  by default the macro searches for ESD in $ALICE_ROOT/MUON/test_out.100
  The directory is generated by MUON/AlirootRun_MUONtest.sh (see README in MUON).

  - Grid test (interactive):
  root
  .x AnalysisTrigChEff.C(kMlocalGrid)

  by default the macro searches for a wn.xml in the current directory, with the list
  of grid files to analyse.

- Display results:
The macro creates the file MUON.TriggerEfficiencyMap.root
aliroot
AliMUONTriggerEfficiencyCells effCells("MUON.TriggerEfficiencyMap.root")
effCells.DisplayEfficiency()

===================================================
 Analysis task for checking the consistency of ESD and AOD filling and contents.
 Mercedes LOPEZ NORIEGA, Orsay
----------------------------------------------------------------

Purpose:check the consistency between the ESDs and the AOD.

The task works locally and on the grid, there are two macros depending where
you want to run

Files:
AliAnalysisTaskAODvsESD.cxx   -source files of the analysis task
AliAnalysisTaskAODvsESD.h

AnalysisTrainMuonComp.C       - macro for analysis on the grid
AnalysisTrainMuonCompLocal.C  - macro for local analysis

To run the check:
1. Copy the following packages in the current working directory
  - STEERBase.par
  - ESD.par
  - AOD.par
  - ANALYSIS.par
  - ANALYSISalice.par
  - PWG3muon.par

2. To run locally:
  root
  .x AnalysisTrainMuonCompLocal.C

  by default the macro searches for a tag file of ESDs in the 
  working directory

3. To run on the grid:
  root
  .x AnalysisTrainMuonComp.C

  by default the macro searches for a wn.xml in the current directory, with the list
  of grid files to analyse.

4. Display results:
The macro creates the file AODvsESDoutput.root
It contains an Ntuple from the ESDs and Ntuple from the AODs and an inv. mass histo 
for each case

===================================================
 Analysis task for extracting variables from the ESD for single muons
 Nicole BASTID and Bogdan VULPESCU, Clermont-Ferrand
----------------------------------------------------------------

Purpose: the output is an ntuple with single muon tracks selected variables; a
histogram collects the number of events in different trigger classes.

Files:

AliAnalysisTaskSingleMuESD.cxx  - source files of the analysis task
AliAnalysisTaskSingleMuESD.h

RunSingleMuonAnalysisFromESD.C  - macro to run the analysis (with xml input
                                  collection and grid connection)


How to run:

1. Copy the following packages in the current working directory
  - STEERBase.par
  - ESD.par
  - AOD.par
  - ANALYSIS.par
  - ANALYSISalice.par
  - PWG3muon.par

2. Run root

gROOT->LoadMacro("$ALICE_ROOT/PWG3/muon/RunAnalysis.C");
setupPar("STEERBase"); 
setupPar("ESD"); 
setupPar("AOD"); 
setupPar("ANALYSIS"); 
setupPar("ANALYSISalice"); 
setupPar("PWG3muon");

3. To analyse grid files, provide an XML collection file wn.xml and run

gROOT->LoadMacro("$ALICE_ROOT/PWG3/muon/RunSingleMuonAnalysisFromESD.C");
RunSingleMuonAnalysisFromESD()

4. To analyse local files, modify RunSingleMuonAnalysisFromESD.C

  } else {
    chain->Add("/path_1_to/AliESDs.root");
    chain->Add("/path_2_to/AliESDs.root");
    chain->Add("/path_3_to/AliESDs.root");
  }

and run (notice the "1" argument)

gROOT->LoadMacro("$ALICE_ROOT/PWG3/muon/RunSingleMuonAnalysisFromESD.C");
RunSingleMuonAnalysisFromESD(1)

5. Check the results in SingleMuESD.root

Note: for some older runs (PDC08/LHC08t for instance) it is necessary to force 
the values of the trigger mask, because the function 

GetFiredTriggerClasses()

from AliESDEvent does not return the proper names of the muon trigger classes.
So one has to un-comment the line

task->SetTriggerType("MUON");

in the macro RunSingleMuonAnalysisFromESD.C, if the simulation was produced 
with SetMakeTrigger("MUON")

This bug is fixed starting with the TRUNK from 14.11.2008

===================================================
 Analysis task for RecoCheck analysis of ESD and Kinematics
 Bogdan VULPESCU, Clermont-Ferrand
----------------------------------------------------------------

Purpose: the class MUON/AliMUONRecoCheck is used to extract single muon and
muon pair information in objects of type AliMUONTrackLight and 
AliMUONPairLight. The output file is RecoCheck.root and contains a tree with
arrays of both objects.

Files:

AliAnalysisTaskRecoCheck.cxx  - source files of the analysis task
AliAnalysisTaskRecoCheck.h

RunRecoCheck.C  - macro to run the analysis (with xml input
                  collection and grid connection)


How to run:

1. Run aliroot (not root!, the class is in the muondep library)

2. gROOT->LoadMacro("$ALICE_ROOT/PWG3/muondep/RunRecoCheck.C");

3. RunRecoCheck()

4. To analyse local files, modify RunRecoCheck.C

  } else {
    chain->Add("/path_1_to/AliESDs.root");
    chain->Add("/path_2_to/AliESDs.root");
    chain->Add("/path_3_to/AliESDs.root");
  }

and run (notice the "1" argument)

RunRecoCheck(1)

5. Check results in RecoCheck.root

Note: the magnetic field is needed in order to initialize the tracker. The
corresponding field map is selected by looking at the value of the L3 current
stored in the GRP database which was used for the reconstruction.

===================================================
 Analysis task to correlate ESD tracks to their corresponding Monte Carlo data.
 Artur Szostak, INFN
---------------------------------------------------------------

The AliAnalysisTaskLinkToMC analysis task is used to correlate reconstructed tracks
found in ESDs to their corresponding MC track information found in the kinematics
tree. The the same algorithm is used as in MUON/AliMUONRecoCheck but without any
dependancy on MUON library code. This analysis task will copy out ESD MUON track
information into the AOD like AliAnalysisTaskESDMuonFilter does, however the MC
track label in the AOD track class is filled with the label of the corresponding
MC track. Thus, subsequent analysis tasks can easily find the correct MC track from
the label.

Files:

AliAnalysisTaskLinkToMC.h    - The header file for the analysis task code.
AliAnalysisTaskLinkToMC.cxx  - The source file for the analysis task code.

RunLinkToMCAnalysisExample.C - An example macro for running the analysis task in
  local mode over simulated data. The macro takes one argument which is the name
  of the ESD file to process. By default it uses AliESDs.root in the current working
  directory.

PlotEfficiency.C  - This macro can be used to quickly check the control histograms
  which are generated by RunLinkToMCAnalysisExample.C by default. The macro simply
  calculates the reconstruction efficiency from the found and findable track histograms
  that are filled by the analysis task.

How to run:

To run the analysis task in local mode with the example macro simply run the command
 $ root RunLinkToMCAnalysisExample.C\(\"AliESDs.root\"\)
where AliESDs.root can be replaced with the full path and name of the appropriate ESD
file you wish to process. This command can also be run via AliRoot.


===================================================
 Analysis task to create mixed AOD events, containing a 
 pair of muon tracks coming from two different real AOD events.
 Alessandro De Falco and Antonio Uras, INFN Cagliari 
---------------------------------------------------------------

IMPORTANT: The AliAnalysisTaskCreateMixedDimuons analysis task needs
           the class AliEventPoolMuon.

In class AliEventPoolMuon the definition of the pools to be used 
in the mixing is implemented. Up to now, it is possible to create 
pools diving events by:
 - total multiplicity
 - forward muons multiplicity
 - z-position of the primary vertex

In class AliAnalysisTaskCreateMixedDimuons the mixing is performed between
each pair of events which can be built up picking events from the internal 
buffer: if the buffer contains n events, n(n-1)/2 mixed events are created 
within the same call of the task.

Some more details:
 - the mixing task asks a completely fresh buffer for each call
 - the mixing task creates output events from scratch
 - output events are filled with two muon tracks (picked randomly
   from the events being mixed and a vertex having its z-position in the middle
   of the z-vertex bin defining the pool. Due to technical reasons,
   the tracks are NOT connected with the vertex
 - standard AOD are created as output, nearly completely empty

Files:

AliAnalysisTaskCreateMixedDimuons.(h,cxx)  - The analysis task class.
AliEventPoolMuon.(h,cxx)                   - The class implementing the pool.

AnalysisTrainCreateMixedDimuons.C - An example macro enabling connections between
the pool manager, the input event handler, the input and output containers, 
via the Analysis Manager. Input arguments: - directory containing AOD tag file
                                           - name of the output file
                                           - number of events to be loaded
                                             in the internal buffer

How to run:

To run the analysis task in local mode with the example macro, simply run
the AnalysisTrainCreateMixedDimuons.C macro (without compiling it) with 
appropriate arguments.

===============================================================================
Analysis Task for (di)muons continuum from heavy flavors
X. M. Zhang, Clermont-Ferrand and Wuhan
===============================================================================
1) AliAnalysisTaskSEMuonsHF.h (.cxx) -- analysis class for (di)muons continuum
   from heavy flavors.
2) AliMuonsHFHeader.h (.cxx) -- class used to collect and save the information
   at event level (eg. vertex position and trigger mask).
3) AliAODMuonTrack.h (.cxx) -- class used to extract and save information of
   reconstracted muon tracks (from both AOD and ESD), selection the muon with
   the set cuts.
4) AliAODMuonPair.h (.cxx) -- class used to extract and save information of
   muon pairs (both for like-sign and unlike-sign), the corresponding two
   single muon tracks are linked by TRef.
5) AliMCMuonTrack.h (.cxx) -- inherit from AliAODMuonTrack, class used to
   extract and save the Monte-Carlo information of reconstracted muon tracks,
   the generated kinematics and the history of the muon are filled in this
   class.
6) AliMCMuonPair.h (.cxx) -- inherit from AliAODMuonPair, class used to
   extract and save the information of muon pair at Monte-Carlo level, the
   relation between two muons can be obtain from this class.

- the input of AOD and ESD are detected automatically.
- three analysis modes:
  mode0, analyze both single muon and dimuon
  mode1, analyze only the single muon
  mode2, analyze only the dimuon
- two kinds of outputs:
  - histogram, histograms for single muon and/or dimuon will be outputed
    according to different running mode. Events with unreconstructed vertex
    are discarded, cuts are set for (di)muons selection.
  - tree, if flag for output tree switch on, tree with braches of event header
    and/or single muon and/or dimuon will be gotten according to different
    running mode. All events and all recontructed tracks are used to fill the
    tree.
- running with and without MC information:
  - without MC information, AliAODMuonTrack and/or AliAODMuonPair are/is used
    and filled in the tree (if switch on), output the histograms just with
    reconstructed information.
  - with MC information, AliMCMuonTrack and/or AliMCMuonPair are/is used and
    filled in the tree (if switch on), additional histograms for different
    single muon and/or dimuon source are outputed.
  - at present the methods used to get the MC information from ESD is not
    implemented for they depend on the MUON module. Although, we still keep
    these codes in AliMCMuonTrack for private running, more details can be
    found in AliMCMuonTrack.cxx 
- all the flags and cuts are set in AddTaskMuonsHF.C
    Int_t mode = 0;          \\ = 0 analyze both single muon and dimuon
                             \\ = 1 analyze the single muon only
                             \\ = 2 analyze the dimuon only
    Bool_t isMC = kTRUE;     \\ get MC information
    Bool_t isTree = kFALSE;  \\ flag for tree output
    AliAnalysisTaskSEMuonsHF *taskMuonsHF = AddTaskMuonsHF(mode, isMC, isTree);


===================================================
 New task to mesure the spectrometer resolution
 Philippe Pillot, Subatech
---------------------------------------------------------------
1) Compute the cluster resolution per chamber, half chamber and detection element by using 2 methods:
    a) combining the cluster-track residuals obtained with and without using the cluster to reconstruct the track
    b) subtracting the track resolution component to the cluster-track residual obtained without using the cluster to reconstruct the track

- In principle, this task should be run iteratively, using the chamber resolution computed in the step n-1 to weight the clusters when refitting the tracks in the step n.
--> The macro RunMuonResolution.C gives an example on how to do it locally or in PROOF.
Neverthless, the convergence is quite fast so it can also be run only once provided that the starting chamber resolutions are closed to the real one (e.g. to monitor the variation of the resolution with time).

- If not specified by user when initializing the task, the starting chamber resolution is taken from the recoParam. In addition, the task load several data from the OCDB (mapping, magnetic field...) so the default storage must be properly set.

2) Return the momentum, transverse momentum and angular resolutions of the reconstructed tracks at first cluster and at vertex (i.e. including absorber effects). These resolutions are computed by using the chamber resolutions given as input (i.e. the ones from the step n-1 or the starting ones).

3) Finally, it is possible to re-align the clusters before computing the resolution, by providing the OCDB path to the old and new alignment data (see comments in the code for more details).