Likelihood study
Go back to T2KK page


Index:



Current Work

070117

Precut efficiency with errors (gif) (eps)
070116

I made some plots for some efficiencies:

Precuts efficiency: (gif) (eps)
Precuts efficiency (only nue):
(gif) (eps)

How to run the code

Before to be able to run my likelihood code (run in DASH), I have to make sure that all the right variables are in the hbook files.

This work is done on suketto.

Need to make new zbs files containing
    - Polfit2 variables (need to run polfit2)
                /home/fdufour/work24-t2kk/polfit2
         /home/fdufour/work24-t2kk/mypolfit2

        In the .sh file we must have:

    For SK-I: (as in polfit2.sh)
setenv PACKAGE /home/atmpd/skrep
setenv PACKAGE_LEVEL 02b_sk2final
setenv EXPERIMENT /skam/skrep
setenv EXPERIMENT_LEVEL 01a
setenv CVSCOSRC /home/atmpd/skrep/02b_sk1final/config

   
For SK-II: (as in sk2.sh)
 
setenv PACKAGE /home/atmpd/skrep
setenv PACKAGE_LEVEL 06a
setenv EXPERIMENT /skam/skrep
setenv EXPERIMENT_LEVEL 05a
setenv CVSCOSRC /home/atmpd/skrep/06a/config
unsetenv EXPERIMENT_ROOT
unsetenv PACKAGE_ROOT
source $CVSCOSRC/imake.cshrc


    - Danka's variables (need to add epi0sep bank). This is a DASH code.
                /home/fdufour/work24-t2kk/make_zbs

 Once I have the new zbs files, I need to create new ntuples by running fillnt.F.
                    /home/fdufour/epi0/likelihood/ntuple/ (for the executable)
         /home/fdufour/work24-t2kk/t2kk/  (and work23 for scripts)
   
Only then can I run the likelihood code.

NB    for 100yrs of MC I need ??? of disk space
         for SK-I data I need ??? of disk space  



The rest of the work is run on neutrino.
Likelihood code:

General structure:

Everything is located in: /work/fdufour/t2kk/t2k-like


Note: both parts of the code need precuts.F

Run the "training" code in order to create the tables needed.

    atm-tables.kumac
    atmsk2-tables.kumac
    t2k-tables.kumac

and the main code is ftables.F

Then I have to run the likelihood analysis itself.
        atm-like.kumac

    atmsk2-like.kumac

        t2k-like.kumac

and the main code is likelihood_7.F


Details to run properly:

To decide which variable to use in the likelihood, use the file
variable.txt where:
    1. efrac
    2. totpe
    3.xalong
    4.cosopen


The code is clearly not perfect right now and some paramters are hard-coded. So be careful to recompile the code, with the
parameters you want depeding on which normalization/MC sample you want.

* Set SK version
      sk_version=1
      write(*,*) 'SK version =', sk_version

If running for SK MC, then choose between SK1 and SK2, if running T2K MC then chose sk_version=1


* Set normalization code:
      code=1

********************************************************
 Assign weighting factor depending on which part of the code I use:
     Training=1
     Running=2
     atm or count=3
     signal nu_e =4
     simulate off-axis angle=5
********************************************************





   




back to top
Official Plots & Tables

For plots going up to 10 GeV (only without CCQE and gif available) go there

Without CCQE hatched:
Atmospheric MC SK-I

Go to atmplots
Atmospheric MC SK-II

Go to atmsk2plots
T2K MC

Go to T2K plots
Likelihood (total) (gif) (eps) Likelihood (total) (gif) (eps) Likelihood (total) (gif) (eps)
Likelihood (ebin) (gif) (eps) Likelihood (ebin) (gif) (eps) Likelihood (ebin) (gif) (eps)
nring (gif) (eps) nring (gif) (eps) nring (gif) (eps)
pid (gif) (eps) pid (gif) (eps) pid (gif) (eps)
cosnue (gif) (eps) cosnue (gif) (eps) cosnue (gif) (eps)
pi0 mass (gif) (eps) pi0 mass (gif) (eps) pi0 mass (gif) (eps)
pi0 like (gif) (eps) pi0 like (gif) (eps) pi0 like (gif) (eps)
energy fraction (gif) (eps) energy fraction (gif) (eps) energy fraction (gif) (eps)
totpe (gif) (eps) totpe (gif) (eps) totpe (gif) (eps)
xalong (gif) (eps) xalong (gif) (eps) xalong (gif) (eps)
cos(open) (gif) (eps) cos(open) (gif) (eps) cos(open) (gif) (eps)


With CCQE hatched:
Atmospheric MC SK-I

Go to atmplots
Atmospheric MC SK-II

Go to atmsk2plots
T2K MC

Go to T2K plots
Likelihood (total) (gif) (eps) Likelihood (total) (gif) (eps) Likelihood (total) (gif) (eps)
Likelihood (ebin) (gif) (eps) Likelihood (ebin) (gif) (eps) Likelihood (ebin) (gif) (eps)
nring (gif) (eps) nring (gif) (eps) nring (gif) (eps)
pid (gif) (eps) pid (gif) (eps) pid (gif) (eps)
cosnue (gif) (eps) cosnue (gif) (eps) cosnue (gif) (eps)
pi0 mass (gif) (eps) pi0 mass (gif) (eps) pi0 mass (gif) (eps)
pi0 like (gif) (eps) pi0 like (gif) (eps) pi0 like (gif) (eps)
energy fraction (gif) (eps) energy fraction (gif) (eps) energy fraction (gif) (eps)
totpe (gif) (eps) totpe (gif) (eps) totpe (gif) (eps)
xalong (gif) (eps) xalong (gif) (eps) xalong (gif) (eps)
cos(open) (gif) (eps) cos(open) (gif) (eps) cos(open) (gif) (eps)


Efficiency tables:
   
Atmospheric MC SK-I
Atmospheric MC SK-II T2K MC
5 variables 5 variables
5 + efrac 5 + efrac
5 + totpe 5 + totpe
5 + Xalong 5 + Xalong
5+ cospen 5+ cospen
5+ efrac+xalong+cosopen 5+ efrac+xalong+cosopen 5+ efrac+xalong+cosopen
All All


S over sqrt (B) plots:     gif  or eps
Each points corresponds to:
    5 variables
   5 + efrac
    5+totpe
    5+xalong
    5+cosopen
    5+efrac+xalong+cosopen
    all


Updated plots: 0.35 < E < 0.85 gif or eps
Updated plots: 1.5 < E gif or eps



back to top

Test and Trials

Index:
  •     August 29, 2006 (Different set of variables for different energies)

August 29th 2006:

When expending the likelihood study to higher energies I did a quick test on using a different set of variables
for different energy ranges. Here are some quick results.

The Basic setup seems to be the best.

Basic setup:
same thing for every energy, all variables minus totpe

2 regions setup:
Up to 5 GeV, used all minus totpe.
Above 5 GeV used all variables minus t
otpe, pi0mass, cosnue, pid

High limit setup:
Up to 2 GeV, used all minus totpe.
Between 2 GeV and 4 GeV used all variables minus totpe, pi0mass, cosnue
Above 4 GeV used all variables minus t
otpe, pi0mass, cosnue, pid

High limit setup (with totpe):
Up to 2 GeV, used all variables.
Between 2 GeV and 4 GeV used all variables minus totpe, pi0mass, cosnue
Above 4 GeV used all variables minus t
otpe, pi0mass, cosnue, pid

Low limit setup:
Up to 1.5 GeV, used all minus totpe.
Between 1.5 GeV and 3 GeV used all variables minus totpe, pi0mass, cosnue
Above 3 GeV used all variables minus t
otpe, pi0mass, cosnue, pid