I'm looking at a rank 3 data set, and I notice that the position numbers across the top start at 1 rather than 0, so they don't match the coordinates printed out by h5dump. The other dimensions seem to be 0-based (left side and menu bar)...

Hopefully what I'm describing makes sense

John Knutson wrote:

Iâ€™m

looking at a rank 3 data set, and I notice that the position numbers

across the top start at 1 rather than 0, so they donâ€™t match the

coordinates printed out by h5dump. The other dimensions seem to be

0-based (left side and menu bar)â€¦

Hopefully what Iâ€™m describing makes sense

John,

You are describing the *data sheet* or *page* number

(according the HDFView help). You donâ€™t say what version of HDFView

you are using but this is a known bug and is fixed in Version v2.6b1

where array indexing starts at 0 (much to the chagrin of Fortran

programmers). By the way, the 0-base convention has changed over the

years in HDFView; it used to be 1-based (much to the chagrin of C

programmers).

â€“dan

## Â·Â·Â·

```
-- Daniel Kahn
Science Systems and Applications Inc.
301-867-2162
```

Using 2.6.1, for SunOS, fwiw... I had downloaded the latest available binary.

Daniel Kahn wrote:

## Â·Â·Â·

John Knutson wrote:

I'm looking at a rank 3 data set, and I notice that the position numbers across the top start at 1 rather than 0, so they don't match the coordinates printed out by h5dump. The other dimensions seem to be 0-based (left side and menu bar)...

Hopefully what I'm describing makes sense

John,

You are describing the /data sheet/ or /page/ number (according the HDFView help). You don't say what version of HDFView you are using but this is a known bug and is fixed in Version v2.6b1 where array indexing starts at 0 (much to the chagrin of Fortran programmers). By the way, the 0-base convention has changed over the years in HDFView; it used to be 1-based (much to the chagrin of C programmers).

--dan

--

Daniel Kahn

Science Systems and Applications Inc.

301-867-2162

Â Â ------------------------------------------------------------------------

_______________________________________________

Hdf-forum is for HDF software users discussion.

Hdf-forum@hdfgroup.org

http://mail.hdfgroup.org/mailman/listinfo/hdf-forum_hdfgroup.org

I tried both v2.6.1 and v2.7 snapshot. Both are 0-based.

## Â·Â·Â·

On 9/1/2010 5:28 PM, John Knutson wrote:

Using 2.6.1, for SunOS, fwiw... I had downloaded the latest available binary.

Daniel Kahn wrote:

John Knutson wrote:

I'm looking at a rank 3 data set, and I notice that the position numbers across the top start at 1 rather than 0, so they don't match the coordinates printed out by h5dump. The other dimensions seem to be 0-based (left side and menu bar)...

Hopefully what I'm describing makes sense

John,

You are describing the /data sheet/ or /page/ number (according the HDFView help). You don't say what version of HDFView you are using but this is a known bug and is fixed in Version v2.6b1 where array indexing starts at 0 (much to the chagrin of Fortran programmers). By the way, the 0-base convention has changed over the years in HDFView; it used to be 1-based (much to the chagrin of C programmers).

--dan

--

Daniel Kahn

Science Systems and Applications Inc.

301-867-2162

Â Â ------------------------------------------------------------------------

_______________________________________________

Hdf-forum is for HDF software users discussion.

Hdf-forum@hdfgroup.org

http://mail.hdfgroup.org/mailman/listinfo/hdf-forum_hdfgroup.org

_______________________________________________

Hdf-forum is for HDF software users discussion.

Hdf-forum@hdfgroup.org

http://mail.hdfgroup.org/mailman/listinfo/hdf-forum_hdfgroup.org