Tests and extension on PHOTOS: case of exact first order Matrix element in W --> lepton neutrino (gamma)
by Gizo Nanava and Zbigniew Was
Preliminary presentation of results: status as of 25 July 2008.

In the following 3 points we list results of our work:

A) First comparison of results from PHOTOS MC (exact matrix element included) and SANC,
further plots:
PHOTOS without correcting weight and SANC
PHOTOS with correcting weight as since 2003 and SANC
PHOTOS with exact ME weigt and SANC
gamma mu invariant mass from PHOTOS and SANC: this group of plots is added to confirm results of point B seee just below. Having in mind good performace of PHOTOS once correcting weight of year 2003 is used (confirmed in particular by earlier results of this point) it was somewhat surprizing to see such large differences due to matrix element, at the high end of gamma-mu mass spectrum.
This indicates that validation is never completed process. Differences of this last plots result from small fraction of a whole W leptonic decay sample. This sub-sub-permillee patr of the sample may be important after some selection cuts. Previous plots validate standard PHOTOS for essentially all applications. This last one, points that matrix element correcting weight may be needed for simulation of backgrounds for heavy lepton production, for example.

B) Comparison of results from PHOTOS MC: exact matrix element vs. standard version:
MC tester used

C) routines to be replaced or added.
Warning: you may need to adjust HEPEVT common block size by hand.

D) There were only limited tests performed for program running at higher than first order. We have calculated (with 10^7 samples) that at first order in 93.61 % of cases invariant mass of muon plus neutrino is larger than 98% of the maximum . At fsecond order it is 93.76 % and for exponentiation it is 93.78 % These there results are consistent, within expected accuracy with zero order exponentiation. This provides non-conclusive test, that our correcting weight matches requirements of higher order applications. The conclusive tests require detailed comparison with Monte Carlo program of at least complete double photon emission in W decay.

As one can see agreement is rather encouraging and good.
That is why there is no hurry, or there is no need at all, to incorporate the above special modification to Sstandard version of PHOTOS. We will leave this patch as special code indended for tests only.
It will remain so, unless some feed back will change the opinion.
Then, we will incorporate the presented here code, to the standard mode of PHOTOS operation. This will be done in one of the program future versions. The price is: huge relative increase in program size, some inconvenience for use, incompatible programing styles.

Ad ups for point B input files for MC-tester (root format; there improper description of programs actually used)
input root file for MC-tester (point B); from generation with PHOTOS standard
input root file for MC-tester (point B); from generation with PHOTOS ME


Comment: For definition of plot frames etc. you may need to explore my recent papers on PHOTOS written with P. Golonka or G. Nanava as co-author. Unfortunately, at present, information (and in particular program modification file) presented here, is not for easy/safe use. Particular care and contact with the authors is needed.