BenderTutorialV23 < LHCb

LHCb Web>LHCbComputing>LHCbSoftwareTutorials>BenderTutorial>BenderTutorialV23 (2016-02-06, VanyaBelyaev) (raw view)
---+!! LHCb [[http://cern.ch/lhcb%2Dcomp/Analysis/Bender][Bender]] Tutorial v24r0:  Getting started with [[http://cern.ch/LHCb-release-area/DOC/bender][Bender]]


This hands-on tutorial is an introduction to [[http://cern.ch/LHCb-release-area/DOC/bender][Bender]] - [[http://python.org][Python]]-based user friendly environment for physics analysis.

The purpose of these exercises is to allow you to write a simple algorithm for decay selection and analysis.

The exercises cover the following topics:

%TOC%

If the content of this pseudo-"hands-on" tutorial is a bit boring for you or you know well all the described topics 
please refer to [[http://cern.ch/lhcb%2Dcomp/Analysis/Bender/stul.html][these]] or even
 [[http://www.lib.ru/ILFPETROV/ilf_petrov_12_chairs_engl.txt][these]] pages in between the exercises. 


---++ Prerequisites
It is assumed that you are more or less familiar with the [[LHCbSoftwareTrainingBasics][basic tutorial]], 
also some level of familiarity with the  [[DaVinciTutorial][DaVinci tutorial]] is assumed.  
Some familiarity with basic GaudiPython and python-based 
[[LHCb.FAQ.LHCbFAQ#GaudiHistosPython][histograms]] & [[LHCb.FAQ.LHCbFAQ#GaudiTuplesPython][N-tuples]]  
is welcomed. It is also recommended to follow [[LoKiTutorial][LoKi tutorial]]
and the basic  [[http://cern.ch/lhcb-reconstruction/Python/Dst_as_Ntuple_files/frame.htm][GaudiPython tutorial]]. 
You are also invited to the [[mailto:lhcb-bender@cern.c][lhcb-bender]] mailing list. 

---++ Setup the environment for [[http://cern.ch/LHCb-release-area/DOC/bender][Bender]]
For this particular tutorial we'll concentrate on the interactive jobs and let the batch system and GRID, Ganga and DIRAC tool some rest.
Batch and GRID-aware actions for [[http://cern.ch/LHCb-release-area/DOC/bender][Bender]]-based analysis are not covered by this tutorial.

%SYNTAX{ syntax="sh" numbered="1000" numstep="1"}%
> SetupProject Bender v24r0
%ENDSYNTAX% 

---++ Shortcuts for solutions & examples 
For shortcuts, one can have a look into _solutions_, that are accessible in 
=Tutorial/BenderTutor= package:

%SYNTAX{ syntax="sh" numbered="1000" numstep="1"}%
> ls -al $BENDERTUTORROOT/solutions
%ENDSYNTAX% 

For examples, discussed in this tutorial, see the code in 
%SYNTAX{ syntax="sh" numbered="1000" numstep="1"}%
> ls -al $BENDERTUTORROOT/python/BenderTutor
%ENDSYNTAX% 

---++ Structure of Bender-based python module
The general structure of Bender-based python module consists of three parts:
   0  User code, usually the set of import-statements and algorithms
   0  Configuration part, where one configures the job. Note that Bender does not live in vacuum, and certain configuration of job is need
   0  A few lines that actually start the job

None of these part is actually mandatory, but in case one follows this structure, one gets universal python module that is equally 
well applicable for pure interactive  testing using interactive prompt, for small interactive-like jobs, for batch-like jobs and for Grid.

#DoNothing
---+++ Right structure of Bender module:  simple _"do-nothing"_ module

The simplest  _"do-nothing"_  (even more simple than _"Hello, world!"_), but totally valid module is 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## 1) some user code :
def run ( nEvents ) :
    
    for i in range( 0 , min( nEvents , 10 ) ) :  print ' I run event %i ' % i
        
    return 0

## 2) configuration step 
def configure (  datafiles ,  catalogs = [] , castor = False , params = {} ) :   
    
    print 'I am configuration step!'
    return 0

## 3) steer the job
if '__main__' == __name__ : 
    
    print 'This runs only if module  is used as the script! '
    
    run ( 10 ) 
%ENDSYNTAX%

It is highly desirable and recommended to put some "decorations" a top of this minimalistic lines:
   0 add magic  <code>#!/usr/bin/env python</code> line as the top line of the module/script 
   0 make the script executable: <code>chmod +x ./DoNothing.py</code>
   0 add a python documentation close to the begin of the script
   0 fill some useful python attributes with the proper informaton
      * <code>__author__</code>
      * <code>__date__</code>
      * <code>__version__</code>
   0 do not forget to add documenatio in Doxygen-style and use in  comments following tags 
      *  <code>@file</code>
      *  <code>@author</code>
      *  ... 


(1) and (2) will allow you to use module as executable  script without  specifying <code>python</code> as executable program:

%SYNTAX{ syntax="sh" numbered="1000" numstep="1"}%
> ./DoNothing.py 
%ENDSYNTAX% 

For simple scripts the items (3-4) are likely not nesessary, but ait will be very helpful for you (and your colleagues) to reuse your code lines later.

Properly decorated module is available a  <code>$BENDERTUTORROOT/python/BenderTutor/DoNothing.py</code>
 

*IMPORTANT*   _Right structure_ of Bender module is important for GRID submission:  ==Bender== application in  ==Ganga== expects this structure, and ==Ganga== job fail otherwise
 
#HelloWorld
---++ Another useless, less primitive, but still simple _"Hello, world!"_ example 

In this example we'll use the real algorithm, the real input data, the real DaVinci-like configuration of job (but ignoring the actual content of input data)  

---+++ (1) User code: algorithm, that does nothing, just prints  __'Hello,world!'__  
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
# =============================================================================
## import everything from Bender, including the default "run"-function  
from Bender.Main import *
# =============================================================================
## @class HelloWorld
#  simple Python/Bender class for classical 'Hello,World!' example
class HelloWorld(Algo):   ## IMPORTANT: we inherit from Algo-base class 
    """
    The most trivial algorithm to print 'Hello,world!'
    """
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        """
        The main 'analysis' method
        """
        ## use the native Python print to stdout:
        print       'Hello, world! (using native Python)'
        
        ## use Gaudi-style printout:
        self.Print( 'Hello, World! (using Gaudi)')
        
        return SUCCESS      ## IMPORTANT!!! 
# =============================================================================
%ENDSYNTAX% 

---+++ (2) The job configuration:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
# =============================================================================
## The configuration of the job
def configure ( inputdata        ,    ## the list of input files  
                catalogs = []    ,    ## xml-catalogs (filled by GRID)
                castor   = False ,    ## use the direct access to castor/EOS ? 
                params   = {}    ) :
    
    from Configurables import DaVinci
    ## delegate the actual configurtaion to DaVinci 
    dv = DaVinci ( DataType   = '2012' , InputType  = 'MDST' )
    
    ## define the input data
    setData  ( inputdata , catalogs , castor )  ## inform bender about input files and access to data
    
    ## get/create application manager
    gaudi = appMgr() 
        
    ## (1) create the algorithm. defined above 
    alg = HelloWorld( 'Hello' )
    
    ## (2) replace the list of top level algorithm by
    #     new list, which contains only *THIS* algorithm
    gaudi.setAlgorithms( [ alg ] )
             
    return SUCCESS 
# =============================================================================
%ENDSYNTAX%

#SteeringJob
---+++ Finally: (3) job steering for interactive usage:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
# =============================================================================
## Job steering 
if __name__ == '__main__' :

    ## job configuration, get input data from following BK-path:
    ## BKQuery ( '/LHCb/Collision12/Beam4000GeV-VeloClosed-MagDown/Real Data/Reco14/Stripping20/WGBandQSelection7/90000000/PSIX.MDST'   )
    inputdata = [
        '/lhcb/LHCb/Collision12/PSIX.MDST/00035290/0000/00035290_00000221_1.psix.mdst',
        '/lhcb/LHCb/Collision12/PSIX.MDST/00035290/0000/00035290_00000282_1.psix.mdst',
        '/lhcb/LHCb/Collision12/PSIX.MDST/00035290/0000/00035290_00000238_1.psix.mdst',
        ]
    
    configure( inputdata , castor = True )  ## for interactive usage, allow direct access to CASTOR/EOS 
    
    ## event loop, use default run-function from Bender 
    run(10)
        
# =============================================================================
%ENDSYNTAX% 

Complete example (including the proper decoration and documentatiton lines)
 is available a  <code>$BENDERTUTORROOT/python/BenderTutor/HelloWorld.py</code>

#InspectParticles
---++ The first useful example: inspect some particles at  the given TES 
In this example we'll try to __use__ input data and simply inspect the content of certaint TES location at input &mu;DST.
As an input we'll use  &mu;DST from B&Q WG-selection #7, the path in bookkeeping:  
<code>/LHCb/Collision12/Beam4000GeV-VeloClosed-MagDown/Real Data/Reco14/Stripping20/WGBandQSelection7/90000000/PSIX.MDST</code>.

For this example we can easily reuse the [[#SteeringJob][steering part]] form the previous example, but we need to modify 
both user code and configuration parts.

---+++ (1) User code: algorithm, that inspects the content of input TES location 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
# =============================================================================
## import everything from bender 
from Bender.Main import *
# =============================================================================
## @class InspectParticles
#  Inspect particles from the certain TES location 
#  @author Vanya BELYAEV Ivan.Belyaev@nikhef.nl
class InspectParticles(Algo):
    """
    The trivial algorithm to inspect input TES locations 
    """
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        """
        The main 'analysis' method
        """
        ## get *ALL* particles from the input locations 
        particles = self.select ( 'all' ,  ALL )
        if particles.empty() : return self.Warning('No input particles', SUCCESS )

        ## simple dump of particles
        print 'Found  %d particles ' % len ( particles ) 
        print  particles ## dump them! 

        print ' Loop and print only the decay modes: '
        for b in particles: print b.decay() 

        print ' Loop over particles and print name, mass and pT'
        for b in particles:
            print b.name() , 'mass=%.3f GeV, pT=%.2f GeV ' % ( M( b ) / GeV , PT ( b ) / GeV ) 

        return SUCCESS      ## IMPORTANT!!! 
# =============================================================================
%ENDSYNTAX% 
Here <code>ALL</code> at line 1017 is LoKi-functor that accepts all particles. One can use any LoKi functor, 
or expression that evaluated to boolean value, and only those particles from input locations that satisfy 
these criteria will be selected, e.g.   
   * <code> ( PT > 1 * GeV )</code>
   * <code>  in_range ( 2 , Y , 4.5 ) & ( PT > 1 * GeV ) </code>
   * ... etc..
Also, instead of LoKi-functors, the decay descriptors can be used:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
bmesons = self.select ( 'b' , 'Beauty -> J/psi(1S) K+ K-' ) ## very useful when input locations contain several decays 
for b in bmesons : print b.decay() 

jpsifromb = self.select ( 'psi' , 'Beauty -> ^( J/psi(1S) -> mu+ mu-) K+ K-' )  
for p in jpsifromb : print p.decay() 

kaons = self.select('kaons' , 'Beauty -> J/psi(1S) ^K+ ^K-' ) 
%ENDSYNTAX%

The result of one selection can be subsequencely refined uisng three-argument form of <code>self.select</code> statement:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
kaons = self.select('kaons' , 'Beauty -> J/psi(1S) ^K+ ^K-' ) 
positive = self.select('pos' , kaons , Q > 0 ) 
negative = self.select('neg' , kaons , Q < 0 ) 
%ENDSYNTAX%
Here and above  <code>M</code>, <code>PT</code>, <code>Y</code> and <code>Q</code> are  
LoKi-functions that act on the particles. 
See the incomplete list of them [[LoKiHybridFilters][here]] and [[LoKiParticleFunctions][here]].

---+++ (2) The job configuration:
Since in this examepl we do care abotu the input data, the configuration is a little bit more complicated:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## The configuration of the job
def configure ( inputdata        ,    ## the list of input files  
                catalogs = []    ,    ## xml-catalogs (filled by GRID)
                castor   = False ,    ## use the direct access to castor/EOS ? 
                params   = {}    ) :
  
      
    ## import DaVinci 
    from Configurables import DaVinci
    ## delegate the actual configuration to DaVinci
    rootInTES = '/Event/PSIX'
    dv = DaVinci ( DataType   = '2012'    ,
                   InputType  = 'MDST'    ,
                   RootInTES  = rootInTES )
    
    ## add the name of Bender algorithm into User sequence sequence 
    alg_name = 'Inspector'
    dv.UserAlgorithms += [ alg_name ]
    
    ## define the input data
    setData  ( inputdata , catalogs , castor )
    
    ## get/create application manager
    gaudi = appMgr() 
    
    ## (1) create the algorithm with given name 
    alg = InspectParticles (
        alg_name ,
        RootInTES = rootInTES , ## we are reading uDST
        Inputs    = [ 'Phys/SelPsi2KForPsiX/Particles' ]
        )
             
    return SUCCESS 
# =============================================================================
%ENDSYNTAX%

---+++ (3) job steering for interactive usage
For this example we can easily reuse the [[#SteeringJob][steering part]] from the previous example. 

 Complete example (including the proper decoration and documentatiton lines)
 is available as <code>$BENDERTUTORROOT/python/BenderTutor/InspectParticles.py</code>

#InteractiveTips1
---+++ Few tips on interactive usage 
for purely interactive usage it is very convinient to start the python with <code>-i</code> key or 
start <code>ipython</code> instead of bare <code>python</code>:
%SYNTAX{ syntax="sh" numbered="1000" numstep="1"}%
python -i ./InspectParticles.py
%ENDSYNTAX%
In this scenario, after running over the specified number of events, <code>python</code> open the interactive 
prompt and the futher commands could be issues interactively, e.g. one can run over more events 
<code>run(100)</code>; run over all events in input files <code>run(-1)</code>, inspect  the  algorithms: 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
>>> gaudi = appMgr()
>>> alg = gaudi.algorithm('TheNameOfMyAlgorithm')
%ENDSYNTAX%
 A lot of functionality available at interactive python prompt, see e.g.  [[GaudiPythonTutorial][here]] and 
[[http://cern.ch/lhcb-reconstruction/Panoramix/Panoramix_intro.pdf][here]]

#HistosAndTuples
---++ Add the histograms and n-tuples 

It is very easy to extend  [[#InspectParticles][the previous example]] and add the _histograms_ and _n-tuples_  
to the code of the user code

---+++ Adding the histograms
Adding this histograms is just a trivial, one just need to insert the following lines inside <code>analyse</code> method: 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## get *ALL* particles from the input locations 
        particles = self.select ( 'allinputs', ALL  )
        if particles.empty() : return self.Warning('No input particles', SUCCESS )

        ## loop over particles and fill few histograms: 
        for p in particles :
            #          what        histo-ID    low-edge  high-edge  #bins 
            self.plot( PT (p)/GeV , 'pt(B)'  , 0        , 20       , 50  ) 
            self.plot( M  (p)/GeV , 'm(B)'   , 5.2      , 5.4      , 40  ) 
            self.plot( M1 (p)/GeV , 'm(psi)' , 3.0      , 3.2      , 20  )
            self.plot( M23(p)/GeV , 'm(KK)'  , 1.0      , 1.040    , 20  )
%ENDSYNTAX%
The histograms are booked automatically. 

---+++ Adding n-tuples 
Adding _n-tuples_ is also very simple, on ejust need to insert the following lines into <code>analyse</code> method:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## 
        tup = self.nTuple('MyTuple')
        for p in particles :

            tup.column_float ( 'mB'   , M   ( p ) / GeV )
            tup.column_float ( 'ptB'  , PT  ( p ) / GeV ) 
            tup.column_float ( 'mPsi' , M1  ( p ) / GeV )
            tup.column_float ( 'mKK'  , M23 ( p ) / GeV ) 

            tup.write() 
%ENDSYNTAX%

To define the output filenames for the files with  _histograms_ and _n-tuples_, the usual DaVinci configuration is used:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## 
    from Configurables import DaVinci
    ## delegate the actual configuration to DaVinci
    rootInTES = '/Event/PSIX'
    dv = DaVinci ( DataType      = '2012'        ,
                   InputType     = 'MDST'        ,
                   RootInTES     = rootInTES     ,
                   HistogramFile = 'Histos.root' , ## IMPORTANT 
                   TupleFile     = 'Tuples.root'   ## IMPORTANT 
                   )
%ENDSYNTAX%

---+++ Tips and tricks for histograms and tuples 

At the interactive prompt one can make a certain inspection of _histograms_ and _n-tuples_:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
>>> gaudi = appMgr()
>>> alg = gaudi.algorithm('HistosAndTuples')
>>> histos = alg.Histos()
>>> for k in histos :  print alg.Histos(key).dump(20,20) ## dump histogram as text
    ...  
>>> alg.HistoPrint =  True ## make a short summary on booked histograms 
HistosAndTuples   SUCCESS List of booked 1D histograms in directory         "HistosAndTuples" :-
 | ID                        |   Title                                       |    #    |     Mean   |    RMS     |  Skewness  |  Kurtosis  |
 | m(B)                      | "m(B)"                                        |   475   |     5.3108 | 0.060107   |   -0.27166 |    -1.2296 |
 | m(KK)                     | "m(KK)"                                       |   475   |     1.0213 | 0.0071546  |    0.51193 |    0.99013 |
 | m(psi)                    | "m(psi)"                                      |   475   |     3.0991 | 0.030669   |   0.024126 |     1.7185 |
 | pt(B)                     | "pt(B)"                                       |   475   |     4.0809 | 3.4154     |     1.8964 |     4.1495 |
>>> alg.NTuplePrint = True ## get a short summary on n-tuples
HistosAndTuples   SUCCESS List of booked N-Tuples in directory "FILE1/HistosAndTuples"
HistosAndTuples   SUCCESS  ID=MyTuple       Title="MyTuple"                                 #items=4  {mB,ptB,mPsi,mKK}
%ENDSYNTAX%

Complete example (including the proper decoration and documentatiton lines)
 is available as <code>$BENDERTUTORROOT/python/BenderTutor/HistosAndTuples.py</code>

---++ More advanced functionality for _n-tuples_

There is a helper module <code>BenderTools.Fill</code> that allows to simplify filling if _n-tuples_ with more or less common information, 
e.g. for various PIDs and "common" kinematics. 

 For a given particle there is a simple way to put into _n-tuple_ 4-momentum
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
tup        =  ... 
particle =  ...
tup.column ( 'p4' , particle.momentum() / GeV ) ## ann 4-components at once 
%ENDSYNTAX%

#BenderToolsFill
---+++ <code>BenderTools.Fill</code> module 
The module <code>BenderTools.Fill</code> provides functionality for _"power fill"_ of the basic information in one go.
Clearly it does not cover all the cases, but helps to fill many basic kinematical quantities.
The module provides many dedicate functions, e.g. 
   * <code>treatKine</code> to fill the basic kinematic in one go   
   * <code>treatKaons</code>, <code>treatMuons</code>, ... etc...  including pion, protons, diphotons, gammas, ....
   * <code>treatTracks</code> to fill infomation about all daughter tracks 
To activate this functions, see <code>$BENDERTUTORROOT/python/BenderTutor/Tuples2.py</code> example.
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
# ============================================================================= 
## import everything from bender 
from   Bender.Main import *
## enhace functionality for n-tuples 
import BenderTools.Fill 
# =============================================================================
## @class Tuples2
#  Enhanced functionality for n-tuples 
#  @author Vanya BELYAEV Ivan.Belyaev@itep.ru
class Tuples2(Algo):
    """
    Inspect particles from the certain TES location and fill histos and tuple
    """
    def initialize ( self ) :

        sc = Algo.initialize ( self ) ## initialize the base class
        if sc.isFailure() : return sc
        
        ## initialize functionality for enhanced n-tuples 
        sc = self.fill_initialize ()     ## IMPORTANT 
        return sc 

    def finalize ( self ) :    
        self.fill_finalize  ()                  ## IMPORTANT 
        return Algo.finalize ( self )
    
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        """
        The main 'analysis' method
        """
        ## get particles from the input locations 
        particles = self.select ( 'bmesons', 'Beauty -> J/psi(1S) K+ K-'  )
        if particles.empty() : return self.Warning('No input particles', SUCCESS )
        
        tup = self.nTuple('MyTuple')
        for p in particles :
            
            psi = p(1) ## the first daughter: J/psi 
            
            ## fill few kinematic variables for particles,
            ## use suffix to mark the variables propertly            
            self.treatKine   ( tup , p   , '_b'   )
            self.treatKine   ( tup , psi , '_psi' )
            
            self.treatKaons  ( tup , p ) ## fill some basic information for all kaons
            self.treatMuons  ( tup , p ) ## fill some basic information for all muons
            self.treatTracks ( tup , p ) ## fill some basic information for all charged tracks

            tup.write() 

        ## 
        return SUCCESS      ## IMPORTANT!!!
%ENDSYNTAX%

---++++ Variables added by <code>treatKine</code> method

| Name   |  LoKi-functor  and/or meaning | Comment |
| <code>pid</code>   | <code>ID  </code>  | |
| <code>pt</code>    | <code>PT/GeV</code>  | |
| <code>m</code>    | <code>M/GeV</code>  | |
| <code>eta</code>   | <code>ETA </code>  | |
| <code>phi</code>   | <code>PHI </code>  | |
| <code>p4(E,X,Y,Z)</code> | 4-momentum in GeV    || 
| <code>c2ip</code> | &chi;<sup>2</sup>(IP)   using <code>BPVIPCHI2()</code>  | currently for all particles, protection to be added soon  | 
| <code>y</code>   | <code>Y</code>  | for _non-basic_ particles only  |
| <code>lv01</code>   | cos&theta;<sup>*</sup> using <code>LV01</code>  | for _non-basic_ particles only  |
| <code>ctau</code>   | c&times;&tau;  using <code>BPVLTIME()*c_light</code>  | for particles with valid end-vertex only    |
| <code>ctau9</code>   | c&times;&tau;  using <code>BPVLTIME(9)*c_light</code>  | for particles with valid end-vertex only    |
| <code>ctau25</code>   | c&times;&tau;  using <code>BPVLTIME(25)*c_light</code>  | for particles with valid end-vertex only    |
| <code>dtf</code>   | &chi;<sup>2</sup>(DTF)/ndf  using <code>DTF_CHI2NDOF(True)</code>  | for particles with valid end-vertex only, likely to be removed soon  | 
| <code>dls</code>   | <code>LoKi.Particles.DecayLengthSignificanceDV()</code>  | for particles with valid end-vertex only  | 
| <code>vchi2</code>   | &chi;<sup>2</sup>(VX)   using <code>VFASPF( VCHI2 ) </code>  | for particles with valid end-vertex only  | 
| <code>vchi2ndf</code>   | &chi;<sup>2</sup>(VX)/ndf  using <code>VFASPF ( VCHI2PDOF)</code>  | for particles with valid end-vertex only  | 

All actual names of variables  are appended with <code>suffix</code> used in <code>treatKine</code> method.


#FillRecSummary
There is an easy possibility to put into _n-tuple_ the basic infomation from <code>LHCb::RecSummary</code> object:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
 ## get Rec-summary from TES
 rc_summary   = self.get('/Event/Rec/Summary').summaryData()
 tup = self.nTuple ( ... ) 
 self.addRecSummary ( tup , rc_summary )  ## put rec-summary information into n-tuple
%ENDSYNTAX%

#FillOdin
Also the information from <code>LHCb::ODIN</code> objects can be added to _n-tuple_ in one go: 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## get ODIN from TES
odin  = self.get_ ( '/Event/DAQ/ODIN'   , True )
tup   = self.nTuple( ... )
tup.column_aux ( odin ) 
%ENDSYNTAX%

#FillMasses
For given mother particles it is very easy to add information about invariant masses of all daughetr combuinations, 
optionally applying constrains (for constraints <code>DecayTreeFitter</code> is used.
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
tup = self.nTuple( ... ) 
for B in particles : 
       self.fillMasses ( tup , B , '' ) ## just fill masses of all combinations  
       self.fillMasses ( tup , B , 'cPV' , True ) ## fill masses after DTF-refit with PV-constraint
       self.fillMasses ( tup , B , 'cPVM' , True , 'J/psi(1S)' ) ## fill masses after DTF-refit with PV-constraint and J/psi-mass consrtraint
%ENDSYNTAX%

#MakeB 
---++ Making the combinatorics

All previous examples were devoted to the looping over the particles, created at earlier steps, 
e.g. at Stripping. But Bender also allows to perform combinatoric and creation of composite particles.
Let's consider the creation of <latex>B^+\rightarrow J/\psi K^+</latex> candidates in Bender  using
<latex>J/\psi</latex> candidates created earlier in Stripping.
 

---+++ (1) User code: algorithm, that makes B<sup>+</sup> from J/&psi; and K<sup>+</sup> 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## @class MakeB 
#  Make combinatorics and composed particle creation in Bender 
#  @author Vanya BELYAEV Ivan.Belyaev@itep.ru
class MakeB(Algo):
    """
    Make combinatorics and composed particle creation in Bender 
    """
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        """
        The main 'analysis' method
        """
        ## get J/psi mesons from the input
        psis  = self.select ( 'psi' , 'J/psi(1S) -> mu+ mu-' )
        if psis.empty() : return self.Warning('No J/psi candidates are found!', SUCCESS )
        
        ## get energetic kaons from the input: note we select both K+ and K-
        kaons = self.select ( 'k'  , ( 'K+' == ABSID ) & ( PT > 1 * GeV ) &  ( PROBNNk > 0.2 ) )
        
        ## prepare useful functors: mass, chi2 and c*tau from DecayTreeFitter 
        mfit  = DTF_FUN  ( M , True , strings('J/psi(1S)') ) ## use PV and J/psi-mass constraint
        c2dtf = DTF_CHI2NDOF ( True , strings('J/psi(1S)') ) ## ditto
        ctau  = DTF_CTAU ( 0 , True , strings('J/psi(1S)') ) ## ditto 
                                
        tup = self.nTuple('MyTuple')

        ## make a loop over J/psi K combinations :
        for b in self.loop( 'psi k' , 'B+' ) :              ## THIS LINE MAKES COMBINATORICS! 
            ## fast evaluation of mass
            m12 = b.mass(1,2) / GeV 
            if not 4.9 < m12  < 5.6 : continue ## skip bad combinations 

            ## check the common vertex
            vchi2 = VCHI2 ( b ) 
            if not 0<= vchi2 < 20   : continue   ## skip large chi2 and fit failures  
            
            ## get the mass of B
            mass = M ( b ) / GeV 
            if not 4.9 < mass < 5.6 : continue ## skip bad combinations 

            psi   = b(1) ## the first daughter 
            kaon  = b(2) ## the second daughter
            if not psi  : continue
            if not kaon : continue
            
            if   Q ( kaon ) < 0 : b.setPID ( 'B-' ) ## readjust PID 
            else                : b.setPID ( 'B+' ) ## readjust PID 
            
            m_fit  = mfit  ( b ) / GeV
            if not  5.0 <= m_fit  <= 5.5 : continue   ## skip bad combinations 
            c_tau  = ctau  ( b )
            if not -1.0 <= c_tau  <= 100 : continue  ## skip bad combinations 
            c2_dtf = c2dtf ( b )  
            if not -1.0 <= c2_dtf <=  10 : continue  ## skip bad combinations 
                        
            tup.column  (  'mB'    , mass   ) ## simple mass 
            tup.column  (  'mBdtf' , m_fit  )   ## mass after DecayTreeFitter 
            tup.column  (  'c2dtf' , c2_dtf )  ## chi2 from DecayTreeFitter 
            tup.column  (  'ctau'  , c_tau  ) ## lifetime (c*tau) from DecayreeFitter 
            
            tup.write()
            
            ## finally, save good candiated! 
            b.save ( 'MyB' )                        

        ## check  all saved B-candidates and make filter-passed decision  
        savedb = self.selected('MyB')
        if 0 < len ( savedb ) : self.setFilterPassed( True ) 
             
        return SUCCESS      ## IMPORTANT!!! 
%ENDSYNTAX% 

---+++ (2) Configuration step to use ==DIMUON.DST== as the input
At configuration step we need to feed the algorithm with previously selected J/&psi;-candidates and kaons.
Considering ==DIMIUON.DST== form the Stripping20 as the input, the possible configuration looks as:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## The configuration of the job
def configure ( inputdata        ,    ## the list of input files  
                catalogs = []    ,    ## xml-catalogs (filled by GRID)
                castor   = False ,    ## use the direct access to castor/EOS ? 
                params   = {}    ) :
    
    ## read only events with Detached J/psi line fired   IMPORTANT!!! 
    Jpsi_location  = '/Event/Dimuon/Phys/FullDSTDiMuonJpsi2MuMuDetachedLine/Particles'
    from PhysConf.Filters import LoKi_Filters
    fltrs = LoKi_Filters ( STRIP_Code = "HLT_PASS_RE('Stripping.*DiMuonJpsi2MuMuDeta.*')"  )
    
    ## import DaVinci    
    from Configurables import DaVinci
    ## delegate the actual configuration to DaVinci
    dv = DaVinci ( EventPreFilters =  fltrs.filters ('Filters') , ## IMPORTANT
                   DataType        = '2012'       ,
                   InputType       = 'DST'        ,
                   TupleFile       = 'MakeB.root' )
    
    ## add the name of Bender algorithm into User sequence sequence 
    alg_name = 'MakeB'
    dv.UserAlgorithms += [ alg_name ]
    
    from StandardParticles import StdLooseKaons 
    kaons = StdLooseKaons.outputLocation()
    
    ## define the input data
    setData  ( inputdata , catalogs , castor )
    
    ## get/create application manager
    gaudi = appMgr() 
    
    ## (1) create the algorithm with given name 
    alg   = MakeB (
        alg_name  ,
        Inputs            = [ Jpsi_location , kaons ] , 
        ParticleCombiners = { '' : 'LoKi::VertexFitter' } )
             
    return SUCCESS 
%ENDSYNTAX%

---+++ (3) Job steering for interactive usage:
The configuration is essentially the same, as for previosu examples: 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## Job steering 
if __name__ == '__main__' :    
    ## job configuration
    ## BKQuery ( '/LHCb/Collision12/Beam4000GeV-VeloClosed-MagDown/Real Data/Reco14/Stripping20/90000000/DIMUON.DST' )
    inputdata = [
        '/lhcb/LHCb/Collision12/DIMUON.DST/00020198/0001/00020198_00012742_1.dimuon.dst',
        '/lhcb/LHCb/Collision12/DIMUON.DST/00020198/0001/00020198_00015767_1.dimuon.dst',
        '/lhcb/LHCb/Collision12/DIMUON.DST/00020198/0000/00020198_00007306_1.dimuon.dst',
        '/lhcb/LHCb/Collision12/DIMUON.DST/00020198/0001/00020198_00016402_1.dimuon.dst',
        ]
    configure( inputdata , castor = True )
    
    ## event loop 
    run(50000)
%ENDSYNTAX%

Complete example (including the proper decoration and documentatiton lines)
 is available as <code>$BENDERTUTORROOT/python/BenderTutor/MakeB.py</code>

#BenderGanga
---++ Bender & Ganga 
To submit  Bender jobs to GRID using Ganga,  very simple script can be used
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
my_area  = '/afs/cern.ch/user/i/ibelyaev/cmtuser'
my_opts  = my_area + "/Bender_v23r2/Phys/VBWork/work/work_Bc2rho/Bc.py"
#
## prepare job :
# 
j = Job (
    #
    name         = 'MyJob',
    #
    application  = Bender (
    version      = 'v23r4'   ,
    module       = my_opts   ## your Bender-based module
    ) ,
    #
    outputfiles  = [ SandboxFile ( '*.root' ) , SandboxFile ( '*.xml'  ) ] ,
    #
    backend      = Dirac () ,
    #
    splitter     = SplitByFiles ( filesPerJob   =     5 ,
                                  bulksubmit    =  True , 
                                  ignoremissing = False ) 
    )

## get data from BK-dbase 
q = BKQuery ( '/LHCb/Collision12/Beam4000GeV-VeloClosed-MagDown/Real Data/Reco14/Stripping20/WGBandQSelection7/90000000/PSIX.MDST'   )
dataset = q.getDataset()

## set input data to job 
j.inputdata = dataset

## submit the job
j.submit()
%ENDSYNTAX% 


Ganga expect that the module provides two functions:
   0 <code>configure</code>, that gets as the first argument the list of input files and xml-catalogs are supplied as the second argument  
   0 <code>run</code>, that gets an integer argument: number of evens to run

 

#AccessToMCtruth
---++ Access to MC-information

The next simple example illustrate the treatment of MC-truth data in Bender.  By MC-truth data we mean <code>LHCb::MCParticle</code> and 
<code>LHCb::MCVertex</code> structures.  E.g. lets create a _n-tuple_  with kinematic of J/&psi; from [[http://svnweb.cern.ch/world/wsvn/lhcb/DBASE/trunk/Gen/DecFiles/dkfiles/Bs_JpsiKKpipi%2Cmm%3DTightCut.dec][<latex>B_s\rightarrow J/\psi K^+K^-  \pi^+ \pi^-</latex>]] decay. 

Theatment of pure  MC-data is very similar to the treatment of reconstructed data. One just  needs to use another 
class and the base class for  user algorithm <code>AlgoMC</code> instead of <code>Algo</code> and import all functionality 
from  <code>Bender.MainMC</code> instead of <code>Bender.Main</code>

---+++ (1) User code: algorithm, that fills _n-tuple_ with J&psi; kinematics from true MC <latex>B_s\rightarrow J/\psi K^+K^-  \pi^+ \pi^-</latex> decay

%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## import everything from bender 
from   Bender.MainMC import *   ## IMPORTANT: note the module name! 
# =============================================================================
## @class MCtruth 
#   Fill n-tuple with some MC-truth information
#  @author Vanya BELYAEV Ivan.Belyaev@itep.ru
class MCtruth(AlgoMC):                        ## <--- Note the base class here
    """
    Make combinatorics and composed particle creation in Bender 
    """
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        """
        The main 'analysis' method
        """
        ## get "all" decays, ignoring possible intermediate resonances and/or additional photons  
        mybs = self.mcselect ( 'mybs' , '[ B_s0 ==> J/psi(1S) K+ K- pi+ pi-]CC' )

        if mybs.empty() :  ## it should ``never'' happen!!!
            self.Warning('No true Bs-decays are found!')
            allb = self.mcselect( 'allmcb' , BEAUTY )
            print allb
            return SUCCESS

        ## 1) get decays to K*K* 
        mybs1 = self.mcselect ( 'mybs1' , '[B_s0       => J/psi(1S) K*(892)0 K*(892)~0 ]CC' )  

        ## 2) get decays to K* K pi
        mybs2 = self.mcselect ( 'mybs2' , '[ [B_s0]cc  => J/psi(1S) K*(892)0 K- pi+    ]CC' )     

        ## 3) get decays to J/psi phi rho 
        mybs3 = self.mcselect ( 'mybs3' , '[B_s0  =>  J/psi(1S) ( phi(1020) => K+ K- ) rho(770)0 ]CC' )
        
        ## 4) get decays to J/psi phi pi pi 
        mybs4 = self.mcselect ( 'mybs4' , '[B_s0  =>  J/psi(1S) ( phi(1020) => K+ K- ) pi+ pi-   ]CC' )

        ## 5) get decays to J/psi K K pi pi 
        mybs5 = self.mcselect ( 'mybs5' , '[B_s0  =>  J/psi(1S) K+ K- pi+ pi-   ]CC' )

        ## 6)  get decays to psi2S phi  
        mybs6 = self.mcselect ( 'mybs6' , '[B_s0  => ( psi(2S) ==> J/psi(1S) pi+ pi- ) ( phi(1020) =>  K+ K- ) ]CC' )

        ## 7) get decays to psi2S K K  
        mybs7 = self.mcselect ( 'mybs7' , '[B_s0  => ( psi(2S) ==> J/psi(1S) pi+ pi- ) K+ K- ]CC' )

        ## 8) get decays to X(3872) phi  
        mybs8 = self.mcselect ( 'mybs8' , '[B_s0  => ( X_1(3872) ==> J/psi(1S) pi+ pi- ) ( phi(1020) => K+ K- ) ]CC' )

        ## 9) get decays to X(3872) K K
        mybs9 = self.mcselect ( 'mybs9' , '[B_s0  => ( X_1(3872) ==> J/psi(1S) pi+ pi- ) K+ K- ]CC' )

        tup1 = self.nTuple ( 'MyTuple' )        
        tup1.column_int ( 'nBs'  , len ( mybs  ) ) 
        tup1.column_int ( 'nBs1' , len ( mybs1 ) ) 
        tup1.column_int ( 'nBs2' , len ( mybs2 ) ) 
        tup1.column_int ( 'nBs3' , len ( mybs3 ) ) 
        tup1.column_int ( 'nBs4' , len ( mybs4 ) ) 
        tup1.column_int ( 'nBs5' , len ( mybs5 ) ) 
        tup1.column_int ( 'nBs6' , len ( mybs6 ) ) 
        tup1.column_int ( 'nBs7' , len ( mybs7 ) ) 
        tup1.column_int ( 'nBs8' , len ( mybs8 ) ) 
        tup1.column_int ( 'nBs9' , len ( mybs9 ) )
        tup1.write()
        
        if len(mybs) != len(mybs1)+len(mybs2)+len(mybs3)+len(mybs4)+len(mybs5)+len(mybs6)+len(mybs7)+len(mybs8)+len(mybs9) :
            self.Warning('Something wrong happens with decay descriptors!') ## it shoduld never happen! 
            print len(mybs), len(mybs1), len(mybs2), len(mybs3) , len(mybs4), len(mybs5) , len(mybs6) , len(mybs7), len(mybs8),len(mybs9) 
            print mybs

        tup = self.nTuple ( 'MyTuple2' )
        psi_selector = LoKi.MCChild.Selector( 'J/psi(1S)' == MCID )  ## helper utility to get daughter J/psi 
        km_selector  = LoKi.MCChild.Selector( 'K-'        == MCID )     ## helper utiliy to get daughter K- 
        kp_selector  = LoKi.MCChild.Selector( 'K+'        == MCID )     ## helper utilitt to get daughter K+
        
        for b in mybs :

            psi = b.child ( psi_selector )  ## get daughter J/psi
            if not psi : continue

            km  = b.child ( km_selector )  ## get daughter K-
            if not km : continue

            kp  = b.child ( kp_selector )  ## get daughter K+ 
            if not kp : continue

            tup.column ( 'psi' , psi.momentum() / GeV )  ## fill n-tuple 
            tup.column ( 'kp'  , kp .momentum() / GeV )
            tup.column ( 'km'  , km .momentum() / GeV )

            tup.write ()

        ## 
        return SUCCESS      ## IMPORTANT!!! 
%ENDSYNTAX%

Note the usage of <code>self.mcselect</code> instead of <code>self.select</code>. The prefix <code>mc</code> denotes the treatment  
of pure MC-data
   
---+++ (2) Configuration step to use MC-files as the input
The actual configuration for pure MC data is even more simple. 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## The configuration of the job
def configure ( inputdata        ,    ## the list of input files  
                catalogs = []    ,    ## xml-catalogs (filled by GRID)
                castor   = False ,    ## use the direct access to castor/EOS ? 
                params   = {}    ) :
    
    ## import DaVinci    
    from Configurables import DaVinci
    ## delegate the actual configuration to DaVinci
    dv = DaVinci ( Simulation      = True                      , ## IMPORTANT 
                   DDDBtag         = 'dddb-20130929-1'         , ## IMPORTANT 
                   CondDBtag       = 'sim-20130522-1-vc-mu100' , ## IMPORTANT 
                   DataType        = '2012'         ,
                   InputType       = 'DST'          ,
                   TupleFile       = 'MCtruth.root' )
    
    ## add the name of Bender algorithm into User sequence sequence 
    alg_name = 'MCtruth'
    dv.UserAlgorithms += [ alg_name ]
    
    ## define the input data
    setData  ( inputdata , catalogs , castor )
    
    ## get/create application manager
    gaudi = appMgr() 
    
    ## (1) create the algorithm with given name 
    alg   = MCtruth( alg_name  , PPMCs =  [] )
    
    return SUCCESS 
%ENDSYNTAX%


---+++ (3) job steering for interactive usage
%SYNTAX{ syntax="python" numbered="1000" numstep="1}%
if __name__ == '__main__' :
    ## job configuration
    ## data from: BKQuery('/MC/2012/Beam4000GeV-2012-MagUp-Nu2.5-Pythia8/Sim08d/Digi13/Trig0x409f0045/Reco14a/Stripping20NoPrescalingFlagged/13246002/ALLSTREAMS.DST')
    inputdata = [
        '/lhcb/MC/2012/ALLSTREAMS.DST/00033494/0000/00033494_00000013_1.allstreams.dst',
        '/lhcb/MC/2012/ALLSTREAMS.DST/00033494/0000/00033494_00000040_1.allstreams.dst',
        '/lhcb/MC/2012/ALLSTREAMS.DST/00033494/0000/00033494_00000022_1.allstreams.dst',
        '/lhcb/MC/2012/ALLSTREAMS.DST/00033494/0000/00033494_00000004_1.allstreams.dst',
        ]
    
    configure( inputdata , castor = True )
    
    ## event loop 
    run(1000 )
%ENDSYNTAX%

Complete example (including the proper decoration and documentatiton lines)
 is available as <code>$BENDERTUTORROOT/python/BenderTutor/MCtruth.py</code>

#AccessToMCmatch
---++ Access to MC-truth matching information the next  example illustrates the usage of _MC-truth matching_  functionality.
 In Bender it is done using the deficated functor <code>MCTRUTH</code>. 
The functor is fed with certain MC-particles <code>LHCb::MCParticle</code> 
(or several MC-particles) and being applied to _reconstrucructed_ particles <code>LHCb::Particle</code>
is returns <code>True</code> or <code>False</code> depending if this _reconstructed_ particle 
gets direct or indirect contribution from given MC particles.
Lets try to use _MC-truth matching_  functionality to study various MC-truth flags.
The example is a "mixture" of  [[#MakeB][two]] [[#AcessToMCtruth][previous]] examples: 
we'll reconstruct <latex>B_s \rightarrow J/\psi K^+ K^- \pi^+ \pi^-</latex> candidates 
and check if they correspond to MC-truth decays.

---+++ (1) User code: algorithm, that reconstructs <latex>B_s\rightarrow J/\psi K^+K^-  \pi^+ \pi^-</latex> decay and checks MC-truth matching
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## import everything from bender 
from   Bender.MainMC import * 
## @class MCmatch 
#  Very simple manipulations with MC-truth matched events 
#  @author Vanya BELYAEV Ivan.Belyaev@itep.ru
class MCmatch(AlgoMC):                        ## <--- Note the base class here
    """
    Make combinatorics and composed particle creation in Bender 
    """
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        """
        The main 'analysis' method
        """
        ## get "all" decays 
        mybs = self.mcselect ( 'mybs' , '[ B_s0 ==> (J/psi(1S)=> mu+ mu-) K+ K- pi+ pi-]CC' )

        if mybs.empty() :
            self.Warning('No true Bs-decays are found!')
            allb = self.mcselect( 'allmcb' , BEAUTY )
            print allb
            return SUCCESS

        ## get psis from MC-decays 
        mypsi = self.mcselect ( 'mypsi' , '[ B_s0 ==> ^(J/psi(1S)=>mu+ mu-)  K+ K- pi+ pi-]CC' )
        ## get pions from MC-decays : note carets 
        mypi  = self.mcselect ( 'mypi' , '[ B_s0 ==> (J/psi(1S)=>mu+ mu-)  K+ K- ^pi+ ^pi-]CC' )
        ## get psis from MC-decays  : note carets 
        myk   = self.mcselect ( 'myk'  , '[ B_s0 ==> (J/psi(1S)=>mu+ mu-) ^K+ ^K- pi+ pi-]CC' )
        
        ## create mc-truth functors 
        trueB   = MCTRUTH ( mybs  , self.mcTruth () ) ## IMPORTANT !!! 
        truePsi = MCTRUTH ( mypsi , self.mcTruth () ) ## IMPORTANT !!! 
        trueK   = MCTRUTH ( myk   , self.mcTruth () ) ## IMPORTANT !!! 
        truePi  = MCTRUTH ( mypi  , self.mcTruth () ) ## IMPORTANT !!! 

        ## jump to reco-world
        reco_psi  = self.select( 'psi', 'J/psi(1S)' ==    ID )
        if reco_psi.empty() : return SUCCESS
        ## selct only MC-truth matched pions:
        reco_pi   = self.select( 'pi' , ( 'pi+'     == ABSID ) & truePi ) ## NB: use only truth matched pions! 
        if reco_pi .empty() : return SUCCESS
        reco_k    = self.select( 'k'  ,   'K+'      == ABSID            ) 
        if reco_pi .empty() : return SUCCESS
        
        ## prepare useful functors:
        mfit  = DTF_FUN  ( M , True , strings('J/psi(1S)') ) ## use PV and J/psi-mass constraint
        c2dtf = DTF_CHI2NDOF ( True , strings('J/psi(1S)') ) ## ditto
        ctau  = DTF_CTAU ( 0 , True , strings('J/psi(1S)') ) ## ditto 
      
        tup = self.nTuple('MyTuple')
        ##                 1  2 3 4  5 
        Bs = self.loop ( 'psi k k pi pi ' , 'B_s0' )
        for b in Bs :

            jpsi = b(1)
            k1   = b(2)
            k2   = b(3)
            pi1  = b(4)
            pi2  = b(5)

            if 0 < Q( k1)*Q( k2) : continue ## skip wrong charge combinations 
            if 0 < Q(pi1)*Q(pi2) : continue ## skip wrong charge combinations  

            m = b.mass() / GeV 
            if not  4.9 <= m     <= 5.6 : continue
            
            vchi2 = VCHI2 ( b )
            if not -1.0 <= vchi2 <= 100 : continue

            m_fit  = mfit  ( b ) / GeV
            if not  5.0 <= m_fit  <= 5.5 : continue 
            c_tau  = ctau  ( b )
            if not -1.0 <= c_tau  <= 100 : continue
            c2_dtf = c2dtf ( b )  
            if not -1.0 <= c2_dtf <=  10 : continue
            
            tup.column  (  'mBdtf' , m_fit  )
            tup.column  (  'c2dtf' , c2_dtf )
            tup.column  (  'ctau'  , c_tau  )
 
            tup.column_float ( 'm'       , M ( b ) / GeV )
            tup.column_float ( 'mpsi'    , M(jpsi) / GeV )
            
            tup.column       ( 'psi'     , jpsi.momentum() / GeV )
            
            tup.column_bool  ( 'trueB'   , trueB   ( b    )   )  
            tup.column_bool  ( 'truePsi' , truePsi ( jpsi )   )
            tup.column_bool  ( 'trueK1'  , trueK   (  k1  )   )
            tup.column_bool  ( 'trueK2'  , trueK   (  k2  )   )
            tup.column_bool  ( 'truePi1' , truePi  ( pi1  )   )
            tup.column_bool  ( 'truePi2' , truePi  ( pi2  )   )

            tup.write ()

        ## 
        return SUCCESS      ## IMPORTANT!!! 
%ENDSYNTAX% 

---+++ (2) Configuration step to use MC-files as the input 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## The configuration of the job
def configure ( inputdata        ,    ## the list of input files  
                catalogs = []    ,    ## xml-catalogs (filled by GRID)
                castor   = False ,    ## use the direct access to castor/EOS ? 
                params   = {}    ) :
    
    ## read only events with Detached J/psi line fired
    Jpsi_location  = '/Event/AllStreams/Phys/FullDSTDiMuonJpsi2MuMuDetachedLine/Particles'
    from PhysConf.Filters import LoKi_Filters
    fltrs = LoKi_Filters ( STRIP_Code = "HLT_PASS_RE('Stripping.*DiMuonJpsi2MuMuDeta.*')"  )
    
    ## import DaVinci    
    from Configurables import DaVinci
    ## delegate the actual configuration to DaVinci
    dv = DaVinci ( EventPreFilters =  fltrs.filters ('Filters') , ## IMPORTANT
                   Simulation      = True                       , ## IMPORTANT  
                   DDDBtag         = 'dddb-20130929-1'          , ## IMPORTANT 
                   CondDBtag       = 'sim-20130522-1-vc-mu100'  , ## IMPORTANT 
                   DataType        = '2012'       ,
                   InputType       = 'DST'        ,
                   PrintFreq       =  10          , 
                   TupleFile       = 'MCmatch.root' )
    
    ## add the name of Bender algorithm into User sequence sequence 
    alg_name = 'MCmatch'
    dv.UserAlgorithms += [ alg_name ]
    
    from StandardParticles import StdLooseKaons 
    kaons = StdLooseKaons.outputLocation()
    from StandardParticles import StdLoosePions 
    pions = StdLoosePions.outputLocation()
    
    ## define the input data
    setData  ( inputdata , catalogs , castor )
    
    ## get/create application manager
    gaudi = appMgr() 
    
    ## (1) create the algorithm with given name 
    alg   = MCmatch (
        alg_name  ,
        Inputs            = [ Jpsi_location , kaons , pions ] , 
        ParticleCombiners = { '' : 'LoKi::VertexFitter' } , 
        PPMCs             = [ "Relations/Rec/ProtoP/Charged"] ## to speedup a bit 
        )
             
    return SUCCESS 
%ENDSYNTAX%

---+++ (3) job steering for interactive usage
The actual steering part is the same as for [[#AccessToMCtruth][previous]] example.

Complete example (including the proper decoration and documentatiton lines)
 is available as <code>$BENDERTUTORROOT/python/BenderTutor/MCmatch.py</code>


#BenderTisTos
---++ Access to Trigger/TISTOS information in Bender
Bender provides the dedicated module <code>BenderTools.TisTos</code> that greatly simplifies the treatment of Trigger/TISTOS-information.
It allows to put into _n-tuple_ the TISTOS information for the given signal candidates and the set of trigger lines to be checked, as well as to 
get the dump of trigger information to pick up the certain trigger lines fore more detailed investigation.

---+++ (1) User code: algorithm, that gets  <latex>B_s \rightarrow J/\psi K^+ K^-</latex> candidates from input tapes, filter them and fill _n-tuple_ with Trigger/TISTOS information
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## enhance functionality for n-tuples 
import BenderTools.Fill 
import BenderTools.TisTos  ##IMPORTANT! 
# =============================================================================
## @class TisTosTuple
#  Add Trigger/TISTOS info into n-tuple
#  @author Vanya BELYAEV Ivan.Belyaev@itep.ru
class TisTosTuple(Algo):
    def initialize ( self ) :

        sc = Algo.initialize ( self ) ## initialize the base class
        if sc.isFailure() : return sc
        
        ## initialize functionality for enhanced n-tuples 
        sc = self.fill_initialize () 
        if sc.isFailure() : return sc
        
        ## container to collect trigger infomration 
        triggers = {}
        triggers ['psi'] = {}
        ## the lines to be investigated in details
        lines    = {}
        lines    [ "psi" ] = {}
        ## six mandatory keys:
        lines    [ "psi" ][   'L0TOS' ] = 'L0(DiMuon|Muon)Decision'
        lines    [ "psi" ][   'L0TIS' ] = 'L0(Hadron|DiMuon|Muon|Electron|Photon)Decision'
        lines    [ "psi" ][ 'Hlt1TOS' ] = 'Hlt1(DiMuon|TrackMuon).*Decision'
        lines    [ "psi" ][ 'Hlt1TIS' ] = 'Hlt1(DiMuon|SingleMuon|Track).*Decision'
        lines    [ "psi" ][ 'Hlt2TOS' ] = 'Hlt2DiMuon.*Decision'
        lines    [ "psi" ][ 'Hlt2TIS' ] = 'Hlt2(Charm|Topo|DiMuon|Single).*Decision'

        sc = self.tisTos_initialize ( triggers , lines )
        if sc.isFailure() : return sc
     
        return SUCCESS
    
    def finalize ( self ) :
        self.tisTos_finalize  ()       ## do not forget to finalize it 
        self.  fill_finalize  ()       ## do not forget to finalize it  
        return Algo.finalize ( self )  ## do not forget to finalize it 
    
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        ## get particles from the input locations 
        particles = self.select ( 'bmesons', 'Beauty -> J/psi(1S) K+ K-'  )
        if particles.empty() : return self.Warning('No input particles', SUCCESS )
        
        ## prepare useful functors:
        mfit  = DTF_FUN  ( M , True , strings('J/psi(1S)') ) ## use PV and J/psi-mass constraint
        c2dtf = DTF_CHI2NDOF ( True , strings('J/psi(1S)') ) ## ditto
        ctau  = DTF_CTAU ( 0 , True , strings('J/psi(1S)') ) ## ditto 
        
        tup = self.nTuple('MyTuple')
        for b in particles :
            
            psi =b(1) ## the first daughter: J/psi 
            
            c2_dtf = c2dtf ( b )  
            if not -1.0 <= c2_dtf <=   5 : continue
            m_fit  = mfit  ( b ) / GeV
            if not  5.2 <= m_fit  <= 5.4 : continue 
            c_tau  = ctau  ( b )
            if not  0.1 <= c_tau  <=  10 : continue
            
            ## fill few kinematic variables for particles,
            self.treatKine   ( tup , b   , '_b'   )
            self.treatKine   ( tup , psi , '_psi' )

            ## collect trigger information for J/psi
            self.decisions   ( psi , self.triggers['psi'] )

            ## fill n-tuple with TISTOS infornation for J/psi
            self.tisTos      (
                psi    , ## particle 
                tup    , ## n-tuple
                'psi_' , ## prefix for variable name in n-tuple
                self.lines['psi'] ## trigger lines to be inspected
                )
            
            tup.write() 

        ## 
        return SUCCESS      ## IMPORTANT!!! 
%ENDSYNTAX%

The method <code>tisTos</code> adds   following (integer) varibales into _n-tuple_ 
   * <code>psi_l0phys</code>
   * <code>psi_l0tis</code>
   * <code>psi_l0tos</code>
   * <code>psi_l1glob</code>
   * <code>psi_l1phys</code>
   * <code>psi_l1tis</code>
   * <code>psi_l1tos</code>
   * <code>psi_l2glob</code>
   * <code>psi_l2phys</code>
   * <code>psi_l2tis</code>
   * <code>psi_l2tos</code>
(Note the leading prefix <code>psi_</code>).These variables encode the information from 
Trigger/TISTOS decicion for the given J/&psi;-candidate and the specified set of lines.
E.g.  expression <code>((psi_l0t0s&2)==2)</code>  evaluates to <code>True</code> for L0-TOS (J/&psi;)-case, and 
 <code>((psi_l2tis&4)==4)</code> evaluates to <code>True</code> for Hlt2-TIS(J/&psi;)-case.
The description of magic numbers (2,4,...) can be found in [[http://svnweb.cern.ch/world/wsvn/lhcb/Phys/trunk/Phys/DaVinciInterfaces/Kernel/ITisTos.h][ITisTos.h]]-file:
| 0 | Unknown  | 
| 1 | TPS   | 
| 2 | TOS: trigger on signal  |
| 3 | TUS  | 
| 4 | TIS  | 
| 6 | TOS&TIS  | 
| 8 | Trigger decision   |

In case one has an allergy on bit-wise operations, everything can be simplified and  a flag <code>verbose=True</code> could be added 
into <code>tisTos</code> method:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## fill n-tuple with TISTOS infornation for J/psi
            self.tisTos      (
                psi    , ## particle 
                tup    , ## n-tuple
                'psi_' , ## prefix for variable name in n-tuple
                self.lines['psi'] ,  ## trigger lines to be inspected
                verbose = True  ## if someone hates bit-wise operations 
                )
%ENDSYNTAX%
In this case, in addition to variables, listed above new boolean variables  
<code>*_tis</code>, <code>*_tos</code>,  <code>*_tus</code>, <code>*_tps</code> and  <code>*_dec</code>
will be added to _n-tuple_, allowing to avoid bit-wise operations at later steps.

 
 ---+++ (2) Configuration step to  read &mu;DST 
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## The configuration of the job
def configure ( inputdata        ,    ## the list of input files  
                catalogs = []    ,    ## xml-catalogs (filled by GRID)
                castor   = False ,    ## use the direct access to castor/EOS ? 
                params   = {}    ) :
    
    ## import DaVinci 
    from Configurables import DaVinci
    ## delegate the actual configuration to DaVinci
    rootInTES = '/Event/PSIX'
    dv = DaVinci ( DataType      = '2012'        ,
                   InputType     = 'MDST'        ,
                   RootInTES     = rootInTES     ,
                   TupleFile     = 'TisTosTuples.root' ## IMPORTANT 
                   )
    
    ## add the name of Bender algorithm into User sequence sequence 
    alg_name = 'TisTos'
    dv.UserAlgorithms += [ alg_name ]
    
    ## define the input data
    setData  ( inputdata , catalogs , castor )
    
    ## get/create application manager
    gaudi = appMgr() 
    
    ## (1) create the algorithm with given name 
    alg   = TisTosTuple (
        alg_name ,
        RootInTES = rootInTES , ## we are reading uDST
        Inputs    = [ 'Phys/SelPsi2KForPsiX/Particles' ]
        )
             
    return SUCCESS 
%ENDSYNTAX% 


---+++ (3) job steering for interactive usage
The actual steering part is identical to [[#SteeringJob][this one]]


Complete example (including the proper decoration and documentatiton lines)
 is available as <code>$BENDERTUTORROOT/python/BenderTutor/TisTosTuple.py</code>


---++ Accessing to <code>HepMC</code> information
Bender is also able to deal with Generator/<code>HepMC</code> information.
E.g. next simple example gets as input  <code>.gen</code>-file with <latex>B_c^+</latex> 
decays and makes a histogram with transverse momentum of signal meson.
Also it finds other long-lived beauty particle in event and prints its decay.

 ---+++ (1) User code: algorithm, that deals with  <code>HepMC</code> data
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## import everything from bender 
from   Bender.MainMC import * 
# =============================================================================
## @class Generator
#  Very simple manipulations with Generator/HepMC information
#  @author Vanya BELYAEV Ivan.Belyaev@itep.ru
class Generator(AlgoMC):                        ## <--- Note the base class here
    """
    Very simple manipulations with Generator/HepMC information
    """
    ## the main 'analysis' method 
    def analyse( self ) :   ## IMPORTANT! 
        """
        The main 'analysis' method
        """

        
        ## get "all" decays 
        mybc = self.gselect ( 'mybc' , 'B_c+' == GABSID )
        if mybc.empty () : 
            return self.Warning('No true Bc are found!' ,SUCCESS )
        
        for b in mybc :
            self.plot( GPT( b ) / GeV , 'pt(Bc)' , 0 , 50 , 50 )
        
        ## get all other beauty long-lived particles
        from PartProp.Nodes import LongLived
        LongLived.validate ( self.ppSvc() ) 
        otherb = self.gselect ( 'otherb' ,
                                GBEAUTY & ( 'B_c+' != GABSID ) & GDECNODE ( LongLived ) )

        for b in otherb :
            print 'Other B:' , b.decay()

        return SUCCESS
 %ENDSYNTAX%


 ---+++ (2) Configuration step to  read <code>.gen</code>-file 
The configuration step is very easy:
%SYNTAX{ syntax="python" numbered="1000" numstep="1"}%
## The configuration of the job
def configure ( inputdata        ,    ## the list of input files  
                catalogs = []    ,    ## xml-catalogs (filled by GRID)
                castor   = False ,    ## use the direct access to castor/EOS ? 
                params   = {}    ) :
    
    ## import DaVinci    
    from Configurables import DaVinci
    ## delegate the actual configuration to DaVinci
    dv = DaVinci ( Simulation      = True    , ## IMPORTANT 
                   DataType        = '2012'  ,
                   InputType       = 'DST'   ) 
    
    ## add the name of Bender algorithm into User sequence sequence 
    alg_name = 'GenTruth'
    dv.UserAlgorithms += [ alg_name ]

    #
    ## specific acion for HepMC-files
    #
    from BenderTools.GenFiles import genAction
    genAction()
    
    ## define the input data
    setData  ( inputdata , catalogs , castor )
    
    ## get/create application manager
    gaudi = appMgr() 
    
    ## (1) create the algorithm with given name 
    alg   = Generator( alg_name  , PPMCs =  [] )
    
    return SUCCESS 
%ENDSYNTAX%
Topic revision: r8 - 2016-02-06 - VanyaBelyaev
LHCb
Public webs
Welcome Guest
- Cern Search
- TWiki Search
- Google Search
LHCb All webs
Copyright &© 2008-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback