bug fix Description: 1) added a more explainative usage message 2) the percent relative error for the integer type (division) was being done using integer arythmetic; use floating point arythmetic instead 3) added a new test for integer percent Solution: Platforms tested: linux (32,64) AIX solaris Misc. update:
12 lines
670 B
Plaintext
12 lines
670 B
Plaintext
#############################
|
|
Expected output for 'h5diff file1.h5 file1.h5 -v -p 0.02 g1/dset5 g1/dset6'
|
|
#############################
|
|
Dataset: </g1/dset5> and </g1/dset6>
|
|
position dset5 dset6 difference relative
|
|
------------------------------------------------------------------------
|
|
[ 1 0 ] 100 103 3 0.030000
|
|
[ 1 1 ] 100 104 4 0.040000
|
|
[ 2 0 ] 100 150 50 0.500000
|
|
[ 2 1 ] 100 200 100 1.000000
|
|
4 differences found
|