RE: known bug in h2root/makeclass ?

Hi,

I just had a look at the original ntuple in atlfast_500.ntup. In PAW I did:

Hi/file 0 atlfast_500.ntup
Nt/print 3333

It tells me that both ACCHW and STATEHW are R*4 So MakeClass is right making them Float_t

Cheers,
Olivier

-----Original Message-----
From: owner-roottalk_at_root.cern.ch [mailto:owner-roottalk_at_root.cern.ch] On Behalf Of Sebastien Binet
Sent: Wednesday, January 26, 2011 11:19 AM To: roottalk_at_root.cern.ch
Subject: [ROOT] known bug in h2root/makeclass ?

hi there,

I am resurecting an old FORTRAN-based analysis and have to deal with good'ol HBOOK ntuple files (spitted out by ATLFAST FORTRAN and HERWIG)

I tried to migrate it to ROOT/C++ but hit what I think is perhaps a bug in h2root or TTree::MakeClass: the type for 2 leaves ended-up wrong.

the common block declarations to read back these 2 interesting leaves are:

      INTEGER INISTHW, STATEHW
      COMMON/INISTHW/INISTHW(5),STATEHW

      INTEGER ACCHW
      COMMON/ACCHW/ACCHW

while they are interpreted like so by MakeClass:
   Float_t         Acchw;
   Float_t         Statehw;

is this a known problem ?

cheers,
sebastien.

PS: in case it might be useful, I put everything overthere: lxplus418:/tmp/roottalk

--
#########################################


# Dr. Sebastien Binet

# Laboratoire de l'Accelerateur Lineaire # Universite Paris-Sud XI #

Batiment 200 # 91898 Orsay #########################################