D Had Archives - 2007 - September - Recent

09/07/2007

Check the PRD table

dhad update prd table

10/22/2007

Re-run the BF fitter using the new KL eff

• Message from Anders:

Hi,

 Steve brought to my attention earlier today something that
Chul Su noticed. In the MC we use a KS->pipi branching
fraction of 68.61%, however, the PDG is now quoting 69.20+-0.05%.

 It was brought up not so long time ago that in the simulation
we use 68.61% and that this is specified in Geant3:
https://hypernews.lepp.cornell.edu/HyperNews/get/CLEOGSimulation/2.html

 I recall checking at that point that this value was what the
PDG was quoting and I thought that we used the 'right' number.
I might have used an old PDG when I checked...

 I think we have to correct this. This should 'only' affect the
branching fractions for the modes with KS, and not affect the N_DD
and the other branching fractions.

 -Anders

• Werner's follow-up message:

It could either be corrected by hand after the fit OR incorporated as a
K0S efficiency correction (like the pi0 efficiency correction) in the BF
fitter itself.

• Inside "DHadAttr.py"

    'K0S' :
    { 'fractional uncertainty' : 0.018  , 'correction factor' : 1        },
    'pi0' :
    { 'fractional uncertainty' : 0.02   , 'correction factor' : 0.961    },

Change to (69.20/68.61 = 1.009):

    'K0S' :
    { 'fractional uncertainty' : 0.018  , 'correction factor' : 1.009    },
    'pi0' :
    { 'fractional uncertainty' : 0.02   , 'correction factor' : 0.961    },

• Run the fitter

dhad bf K0S

• Make the comparison table

dhad table compare bf K0S

Refresh 7.06/tables/index.muse

• compare bf K0S regular

10/23/2007

Check the error calculation method

• Message from David:

I get somewhat different values for the differences between the
Regular and the KS fit data.  My spreadsheet and a *.pdf version are
included as attachments.  The differences occur in the number of
sigmas between Regular and KS.  Could you please take a look at this
to see what is correct?

The file in PDF, XLS.

Here, David's method of calculating Delta/sig_KS is

 B_KS - B_Reg
--------------
    sig_KS

In DHadTable.py (1.182), compare_bf():

       tab.column_append_by_diff_sigma_pct('diff', bf_A, bf_B, rnd = rnd)

tabletools.py -> column_append_by_diff_sigma_pct() -> column_append_by_divide('a/b', a, b)

here "a" is K0S, "b" is regular.

• Give the two digits table:

dhad table compare bf K0S --set rnd=0.01

• Give another table with reversed comparision

dhad table compare bf regular K0S --set rnd=0.01

• Check the calculation of the sigma

tabletools.py -> column_append_by_diff_sigma_pct ->

        self.column_append_by_subtract_number(title, 'a/b', '1',
                                              rnd, factor, 'sigma')


->
        self.column_trim(title, rnd, factor, err_type)

Revise the code of tabletools.py, so that the current result have the same as David's.

• compare bf K0S regular

10/24/2007

Compose message for the HyperNews

Hi,

I've re-run the BF fitter using Werner suggested method to incorporate a K0S correction in the Fitter.

In the previous fitter, the factor was set to be "1". Now, I set it to 1.009 (

 New BR(K0S-> pi pi)
---------------------  = 69.20/68.61 = 1.009
 Old BR(K0S-> pipi )

A comparison table between the new one and the older one has been make at:

The diff is calculated by:

     Center_Value(K0S) - Center_Value(regular)
  ------------------------------------------------  x 100 %
              Center_Value(regular)

The `sigma' in the parentheses is calculated by:

     Center_Value(K0S) - Center_Value(regular)
  ------------------------------------------------
                Sigma(regular)


http://www.lepp.cornell.edu/~xs32/private/DHad/7.06/tables/compare_bf_K0S_regular.html




Xin

done.

Use more digits for the correction factor

 New BR(K0S-> pi pi)
---------------------  = 69.20/68.61 = 1.00859933
 Old BR(K0S-> pipi )

Use 1.00859933 to do the fit

    'K0S' :
    { 'fractional uncertainty' : 0.018  , 'correction factor' : 1.00859933  },
    'pi0' :
    { 'fractional uncertainty' : 0.02   , 'correction factor' : 0.961    },

• Run the fitter

dhad bf K0S

• Make the comparison table

dhad table compare bf K0S --set rnd=0.01

• compare bf K0S regular

Correlation Matrix

dhad table prd_data_correlation_matrix K0S

• prd data correlation matrix K0S

Correlation changed from regular to K0S in htese :

KS pi pi pi0, KS pi : 0.56 -> 0.55 KS pi pi pi , K pi : 0.29 -> 0.30 KS pi pi pi , K pi pi pi : 0.36 -> 0.37

11/01/2007

Update the PRD table

Message from David:

   Anders and I need a few more numbers from the fit with the new KS
   efficiency:

*  A table of branching fractions, branching ratios and N_DDbar with
   statistical and systematic errors.  These are for Table VIII in the
   paper.

*  Tables of DT and ST efficiencies with the new values of KS
   efficiencies.  These are for Tables III, IV, and V in the paper.

*  We would like to see Regular and KS correlation matrices (total
errors only) with 4 figures after the decimal point rather than 2.  We
want to try to understand if the changes in the correlation matrix are
due to rounding or if there were real changes in the second decimal
place.

   Please let me know if these numbers are easy to obtain from your BF
fitter log file, or if you would have difficulty obtaining them.

dhad table prd_data_results K0S

• prd data results K0S

dhad table prd_dt_dz_eff_yield K0S

• prd dt dz eff yield K0S

No K0S, no change.

dhad table prd_dt_dp_eff_yield K0S

• prd dt dp eff yield K0S

dhad table prd_st_eff_yield K0S

• prd st eff yield K0S

dhad table prd_data_correlation_matrix —set rnd=0.0001

• prd data correlation matrix

dhad table prd_data_correlation_matrix K0S —set rnd=0.0001

• prd data correlation matrix K0S

Updated November 1, 2007