Description of the Milano N-tuples and how to use them

Description of the Milano N-tuples and how to use them

Welcome to Milano N-tuple's world

The following is a description of the N-tuples format adopted by the Milano group for the analysis of the two, three and four body decays of the charmed mesons into charged modes.
For every SKIM1 file which is shipped to WideBand by either Vanderbilt or Boulder, there is a process (called bossanova) which transforms the input file from the conventional packing/unpacking to the HBOOK format of an n-tuple (see here for postscript version of the manual).

The bossanova process consists of a set of PERL scripts that take care of running the FORTRAN conversion program (called drve831_master) and managing all the network and database activities to ship the resulting output files to Milano (pretty much the same as the SKIM1 process does). The drve831_master program is a conventional E831 driver code consisting essentially of the three user subroutines usrnit, usranal and usrstop (see below for the source code of each of them).

For every Run ( Example: Run 10987 ) of Supersteam6 the drve831_master program generates 3 files of N-tuples:

the first file is meant to contain two body decays and is named drve831_master_xxxxx.ss6_2.nhis, where xxxxx stands for run number
(Example: drve831_master_10987.ss6_2.nhis), contains the following SEZDEE modes:

Decay:	Sezdee Mode
K^- π⁺	2
π^- π⁺	10
K^- K⁺	11
K^o_s π⁺	48
K K^o_s	92

the second file is meant to contain three body decays and is named drve831_master_xxxxx.ss6_3.nhis, where xxxxx stands for run number
(Example: drve831_master_10987.ss6_3.nhis), contains the following SEZDEE modes:

Decay:	Sezdee Mode
K^- π⁺ π⁺	1
K⁺ π^- π⁺	4
π⁺ π^- π⁺	6
K⁺ K^- π⁺	7
K⁺ K⁺ π^-	8
K⁺ K^- K⁺	9
K^o_s π⁺ π^-	57
K^o_s K⁺ π^-	90
K^o_s K⁺ K^-	91

the third file is meant to contain four body decays and is named drve831_master_xxxxx.ss6_4.nhis, where xxxxx stands for run number
(Example: drve831_master_10987.ss6_4.nhis), contains the following SEZDEE modes:

Decay:	Sezdee Mode
K^- π⁺ π^- π⁺	3
π^- π⁺ π^- π⁺	12
K^- K⁺ π^- π⁺	13
K^- K^- π⁺ π⁺	14
K^- K⁺ K^- π⁺	15
K^o_s π⁺ π^- π⁺	49
K^o_s K⁺ π^- π⁺	93
K^o_s K^- π⁺ π⁺	94
K^o_s K⁺ K^- π⁺	95

N-tuples variables and their meaning :

Variable	Meaning
`IRUNNO`	Run Number
`ISPILLO`	Spill Number
`IEVENTO`	Event Number
`ITRI`	Bit coded trigger mask: ITRI=2**(J-1) where: J=K if trigger condition K is on J=0 if trigger condition K is off
`MODE`	Decay Mode Number (This is exactly the SEZDEE code)
`MULPRIM`	Multiplicity of Primary vertex
`MICTO`	word containing Pass1, Skim and Milano program versions (see usranal_master.sf to check how this variable is encoded)
`LASKIMW`	Bit coded skim word: LASKIMW= 2**(J-1) where: J=K if skimbit(K) is on J=0 if skimbit(K) is off
`ISTACER(n)`	ISTATP code for particle n: this is a coded variable containing also double link and bucket information as well (see usranal_master.sf to check how this variable is encoded)
`ITRAX(n)`	# of hits for PWC daughter track n, negative if it is identified as muon. For K^o_s, `ITRAX=ITYPVEE`, where `ITYPVEE` is the VEE type
`RMASS`	Invariant mass of candidate
`RMERR`	Invariant mass error of candidate
`DMOM`	Candidate momentum
`XPRIM`	X of the Primary vertex
`YPRIM`	Y of the Primary vertex
`ZPRIM`	Z of the Primary vertex
`XSEC`	X of the Secondary vertex
`YSEC`	Y of the Secondary vertex
`ZSEC`	Z of the Secondary vertex
`ZOUT`	Out Of Material detachment
`CLSEC`	Confidence Level of the secondary vertex
`EL`	Decay length of candidate
`ELSIG`	L/Sigma (detachment)
`AISO1`	Isolation 1
`AISO2`	Isolation 2
`AISO3`	Isolation 3
`PPX5(n)`	X momentum for daughter n
`PPY5(n)`	Y momentum for daughter n
`PPZ5(n)`	( Z momentum for daughter n ) times ( daughter n charge)
`CLBESTO(n)`	Confidence Level of the best CTDL hypotesis CLBESTO=-1 if there is no CTDL information available
`WOBBA(4xn)`	CITADL information (wobs) encoded for i-th daughter `Wobba( (i-1)4 + 1 ) = wobs( 1, i )` `Wobba( (i-1)4 + 2 ) = wobs( 2, i )` `Wobba( (i-1)4 + 3 ) = wobs( 3, i )` `Wobba( (i-1)4 + 4 ) = wobs( 4, i )` for i = 1, n Wobba=-1 if there is not CTDL information
`DSTARTA`	( D^* invariant mass) times ( Bachelor Pion sign ) DSTARTA=0 if incompatible with D^* hypotesis

where:

n=2 for two body decays
n=3 for three body decays
n=4 for four body decays

NOTE :

Data have been written out with an implicit L/σ > 3 cut

The source code

Here is the source code of the three user-routines of drve831_master
(the Milano N-tuple files generator)

Skeletons of N-tuples analysis programs

Following are examples of very simple fortran analysis programs.
The way HBOOK works with n-tuples is the following:

Open a n-tuple file for read (in input)
Loop over events (or one candidates in SEZDEE parlance) and map the content of the n-tuple variables into a common block
Make whatever use suits you of these variables

In short, the examples listed below, do the following:

read a list of N-tuple input files by means of a namelist
perform an analysis reading data from the n-tuple
fill HBOOK histograms
save histograms in HBOOK file.

You can find a sample collection of N-tuple's input files from 6 runs in /nfs-eahome/mount/pdini/ntuple/*.nhis on the eb831.fnal.gov machine at WIDE BAND. These N-tuple files are those same ones listed in the ntupla2.nam namelist file.

Click here for two body analysis program and namelist: ntupla2.f ntupla2.nam

Click here for three body analysis program and namelist: ntupla3.f ntupla3.nam

Click here for four body analysis program and namelist: ntupla4.f ntupla4.nam

Click here for an example of linker command file (usage: linko ntupla2) : linko

NOTE : There are different ways to read a Column Wise N-tuple (CWN) with different
HBOOK routines ( see HBOOK manual ).

WARNING: If you run the program you'll get a message like this :

***** ERROR in Errorpar : New N-tuple, this routine works only for old N-tuples : ID= 4

don't worry: it's just a known old HBOOK bug.

Paolo Dini