Dave, Thank you for your report!
I entered an issue into our database. We will try to resolve it by 1.8.11 release. Elena ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Elena Pourmal The HDF Group http://hdfgroup.org 1800 So. Oak St., Suite 203, Champaign IL 61820 217.531.6112 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ On Oct 29, 2012, at 1:18 PM, Dave Wade-Stein wrote: > So it appears this problem with h5diff extends to h5dump as well. > > I have an h5 file, which, when h5dump'ed on Win32 and Linux, produces > different output. > > Incorrect output is shown in bold. > > Linux: > [dws@oxygen02 history]$ h5dump trajectory > ptclTrackTagFromAbsorber2s_History.h5 > ... > DATASET "trajectory" { > DATATYPE H5T_IEEE_F64LE > DATASPACE SIMPLE { ( 218, 12, 7 ) / ( H5S_UNLIMITED, H5S_UNLIMITED, > H5S_UNLIMITED ) } > DATA { > (0,0,0): 1.06204e-05, -4.79586e-07, 1.00833e+12, 1.87949e+08, 0, 8, 0, > (0,1,0): 0, 0, 0, 0, 0, 0, 0, > (0,2,0): 0, 0, 0, 0, 0, 0, 0, > (0,3,0): 0, 0, 0, 0, 0, 0, 0, > (0,4,0): 0, 0, 0, 0, 0, 0, 0, > (0,5,0): 0, 0, 0, 0, 0, 0, 0, > (0,6,0): 0, 0, 0, 0, 0, 0, 0, > (0,7,0): 0, 0, 0, 0, 0, 0, 0, > (0,8,0): 0, 0, 0, 0, 0, 0, 0, > (0,9,0): 0, 0, 0, 0, 0, 0, 0, > (0,10,0): 0, 0, 0, 0, 0, 0, 0, > (0,11,0): 0, 0, 0, 0, 0, 0, 0, > (1,0,0): 1.07179e-05, -4.79568e-07, 1.01124e+12, 1.89595e+08, 0, 8, 0, > (1,1,0): 0, 0, 0, 0, 0, 0, 0, > (1,2,0): 0, 0, 0, 0, 0, 0, 0, > (1,3,0): 0, 0, 0, 0, 0, 0, 0, > (1,4,0): 0, 0, 0, 0, 0, 0, 0, > (1,5,0): 0, 0, 0, 0, 0, 0, 0, > (1,6,0): 0, 0, 0, 0, 0, 0, 0, > (1,7,0): 0, 0, 0, 0, 0, 0, 0, > (1,8,0): 0, 0, 0, 0, 0, 0, 0, > (1,9,0): 0, 0, 0, 0, 0, 0, 0, > (1,10,0): 0, 0, 0, 0, 0, 0, 0, > (1,11,0): 0, 0, 0, 0, 0, 0, 0, > ... > > Win32: > ... > DATASET "trajectory" { > DATATYPE H5T_IEEE_F64LE > DATASPACE SIMPLE { ( 218, 12, 7 ) / ( H5S_UNLIMITED, H5S_UNLIMITED, > H5S_UNLIMITED ) } > DATA { > (0,0,0): 1.06204e-005, -4.79586e-007, 1.00833e+012, 1.87949e+008, 0, 8, > (0,0,6): 0, > (0,1,0): 1.71412e+214, 1.73429e+185, 1.75754e+136, 3.44266e+218, > (0,1,4): 5.73569e+169, 0, 1.34682e-308, > (0,2,0): 1.91002e-308, 0, 0, 0, 0, 0, 0, > (0,3,0): 0, 0, 0, 0, 0, 0, 0, > (0,4,0): 0, 0, 0, 0, 0, 0, 0, > (0,5,0): 0, 0, 0, 0, 0, 0, 0, > (0,6,0): 0, 0, 0, 0, 0, 0, 0, > (0,7,0): 0, 0, 0, 0, 0, 0, 0, > (0,8,0): 0, 0, 0, 0, 0, 0, 0, > (0,9,0): 0, 0, 0, 0, 0, 0, 0, > (0,10,0): 0, 0, 0, 0, 0, 0, 0, > (0,11,0): 0, 0, 0, 0, 0, 0, 0, > ... > > Oddly, h5dump produces still different output with the -d option. Here is the > output of h5dump -d trajectory, run on Win32: > > HDF5 "ptclTrackTagFromAbsorber2s_History.h5" { > DATASET "trajectory" { > DATATYPE H5T_IEEE_F64LE > DATASPACE SIMPLE { ( 218, 12, 7 ) / ( H5S_UNLIMITED, H5S_UNLIMITED, > H5S_UNLIMITED ) } > DATA { > (0,0,0): 1.06204e-005, -4.79586e-007, 1.00833e+012, 1.87949e+008, 0, 8, 0, > (0,1,0): 0, 0, 0, 0, 0, 0, 0, > (0,2,0): 0, 0, 0, 0, 0, 0, 0, > (0,3,0): 0, 0, 0, 0, 0, 0, 0, > (0,4,0): 0, 0, 0, 0, 0, 0, 0, > (0,5,0): 0, 0, 0, 0, 0, 0, 0, > (0,6,0): 0, 0, 0, 0, 0, 0, 0, > (0,7,0): 0, 0, 0, 0, 0, 0, 0, > (0,8,0): 0, 0, 0, 0, 0, 0, 0, > (0,9,0): 0, 0, 0, 0, 0, 0, 0, > (0,10,0): 0, 0, 0, 0, 0, 0, 0, > (0,11,0): 0, 0, 0, 0, 0, 0, 0, > (1,0,0): 1.07179e-005, -4.79568e-007, 1.01124e+012, 1.89595e+008, 0, 8, 0, > (1,1,0): 0, 0, 0, 0, 0, 0, 0, > (1,2,0): 0, 0, 0, 0, 0, 0, 0, > (1,3,0): 1.37758e-308, 8.34708e-309, 4.17213e-308, 1.043e-308, 0, 0, 0, > (1,4,0): 0, 0, 0, 0, 0, 0, 0, > (1,5,0): 0, 0, 0, 0, 0, 0, 0, > (1,6,0): 0, 0, 0, 0, 0, 0, 0, > (1,7,0): 0, 0, 0, 0, 0, 0, 0, > (1,8,0): 0, 0, 0, 0, 0, 0, 0, > (1,9,0): 0, 0, 0, 0, 0, 0, 0, > (1,10,0): 0, 0, 0, 0, 0, 0, 0, > (1,11,0): 0, 0, 0, 0, 0, 0, 0, > > I've tried versions 1.8.5 through 1.8.9, and they are all wrong, but in > different ways! > > Here is the output from 1.8.5-patch1, run on the same file on Win32: > > DATASET "trajectory" { > DATATYPE H5T_IEEE_F64LE > DATASPACE SIMPLE { ( 218, 12, 7 ) / ( H5S_UNLIMITED, H5S_UNLIMITED, > H5S_UNLIMITED ) } > DATA { > (0,0,0): 1.06204e-005, -4.79586e-007, 1.00833e+012, 1.87949e+008, 0, 8, > (0,0,6): 0, > (0,1,0): 0, 0, 0, 0, 0, 0, 0, > (0,2,0): 0, 0, 1.32116e-308, 1.38263e-304, 0, 0, 0, > (0,3,0): 0, 0, 0, 0, 0, 0, 0, > (0,4,0): 0, 0, 0, 0, 0, 0, 0, > (0,5,0): 0, 0, 0, 0, 0, 0, 0, > (0,6,0): 0, 0, 0, 0, 0, 0, 0, > (0,7,0): 0, 0, 0, 0, 0, 0, 0, > (0,8,0): 0, 0, 0, 0, 0, 0, 0, > (0,9,0): 0, 0, 0, 0, 0, 0, 0, > (0,10,0): 0, 0, 1.32228e-308, 1.38263e-304, 0, 0, 0, > (0,11,0): 0, 0, 0, 0, 0, 0, 0, > > This is an anomalous case for us, as it only appears on this one file, out of > 6200+ in our test suite. We can work around it for now, but it would be great > to know what it is we are doing when we create this file that is causing the > problem. It's small enough (144K) that I figured I'd just attached it so that > anyone who wants to look at the file may do so. > > Thanks! > Dave > <ptclTrackTagFromAbsorber2s_History.h5> > >> I'm getting some output from h5diff which seems wrong. >> >> [dws@oxygen02 history]$ /usr/local/contrib/hdf5-1.8.8-par/bin/h5diff -v -p >> 1e-30 ptclTrackTagFromAbsorber1s_HistoryOld.h5 >> /scr1_oxygen/dws/jenkins/workspace/vorpalall-build/qarnodes/oxygen/vptests/vpresults-oxygen/history/ptclTrackTagFromAbsorber1s/ptclTrackTagFromAbsorber1s_HistoryOld.h5 >> /trajectory >> dataset: </trajectory> and </trajectory> >> size: [339x7x6] [339x7x6] >> position trajectory trajectory difference relative >> >> ------------------------------------------------------------------------ >> [ 0 1 0 ] 1.678294506e+243 0 1.678294506e+243 1 >> >> [ 0 1 1 ] 9.047791808e+164 0 9.047791808e+164 1 >> >> [ 0 1 2 ] 4.78541875e+175 0 4.78541875e+175 1 >> >> [ 0 1 3 ] 4.440349806e-38 0 4.440349806e-38 1 >> >> [ 0 1 4 ] 4.545801862e+97 0 4.545801862e+97 1 >> >> [ 0 1 5 ] 8.535925984e-307 0 8.535925984e-307 1 >> >> [ 0 2 0 ] 3.952525167e-322 0 3.952525167e-322 1 >> >> [ 0 2 1 ] 3.119382268e-319 0 3.119382268e-319 1 >> >> [ 0 2 2 ] 9.698666729e-317 0 9.698666729e-317 1 >> >> [ 0 2 3 ] 6.920039978e-310 0 6.920039978e-310 1 >> >> [ 0 2 4 ] 4.193339982e+228 0 4.193339982e+228 1 >> >> [ 0 2 5 ] 1.129002081e+277 0 1.129002081e+277 1 >> >> [ 0 3 0 ] 1.355073239e+248 0 1.355073239e+248 1 >> >> [ 0 3 2 ] 3.162020133e-322 0 3.162020133e-322 1 >> >> [ 0 3 3 ] 3.162020133e-322 0 3.162020133e-322 1 >> >> [ 0 3 5 ] 2.60988666e+180 0 2.60988666e+180 1 >> >> [ 0 4 0 ] 4.193339982e+228 0 4.193339982e+228 1 >> >> [ 0 4 1 ] 1.129002081e+277 0 1.129002081e+277 1 >> >> [ 0 4 2 ] 1.355073239e+248 0 1.355073239e+248 1 >> >> [ 0 4 5 ] 3.113058228e-319 0 3.113058228e-319 1 >> >> [ 0 5 0 ] 4.824065618e+228 0 4.824065618e+228 1 >> >> [ 0 5 1 ] 2.60988666e+180 0 2.60988666e+180 1 >> >> [ 0 5 2 ] 4.193339982e+228 0 4.193339982e+228 1 >> >> [ 0 5 3 ] 1.129002081e+277 0 1.129002081e+277 1 >> >> [ 0 5 4 ] 1.355073239e+248 0 1.355073239e+248 1 >> >> >> The diffs says that element (0,1,0) of the /trajectory dataset is >> 1.678294506e+243 (in bold above), but when I use h5dump to look at the >> dataset, I see 0s: >> >> ... >> DATASET "trajectory" { >> DATATYPE H5T_IEEE_F64LE >> DATASPACE SIMPLE { ( 339, 7, 6 ) / ( H5S_UNLIMITED, H5S_UNLIMITED, >> H5S_UNLIMITED ) } >> DATA { >> (0,0,0): 1.09772e-05, 1.007e+12, 0, 0, 8, 0, >> (0,1,0): 0, 0, 0, 0, 0, 0, >> (0,2,0): 0, 0, 0, 0, 0, 0, >> (0,3,0): 0, 0, 0, 0, 0, 0, >> (0,4,0): 0, 0, 0, 0, 0, 0, >> (0,5,0): 0, 0, 0, 0, 0, 0, >> (0,6,0): 0, 0, 0, 0, 0, 0, >> (1,0,0): 1.10752e-05, 1.00993e+12, 0, 0, 8, 0, >> (1,1,0): 0, 0, 0, 0, 0, 0, >> (1,2,0): 0, 0, 0, 0, 0, 0, >> (1,3,0): 0, 0, 0, 0, 0, 0, >> (1,4,0): 0, 0, 0, 0, 0, 0, >> (1,5,0): 0, 0, 0, 0, 0, 0, >> (1,6,0): 0, 0, 0, 0, 0, 0, >> ... >> >> Any ideas what's going on here. The datatype in both files is >> H5T_IEEE_F64LE, so I know this is not a datatype comparison issue. >> >> Thanks, >> Dave > > > > > > _______________________________________________ > Hdf-forum is for HDF software users discussion. > [email protected] > http://mail.hdfgroup.org/mailman/listinfo/hdf-forum_hdfgroup.org
_______________________________________________ Hdf-forum is for HDF software users discussion. [email protected] http://mail.hdfgroup.org/mailman/listinfo/hdf-forum_hdfgroup.org
