Datasets are the lowest level.

Tables build on datasets and have an API to work with them.  Still just a
dataset as an implementation.

PacketTables are an optimized Table api designed for frequent appending
such as in a real time application.  The implementation trick they use is
they simply keep the table api opened underneath so each append only
changes the dataset size and then writes.  I like 'em for most applications
as you're no worse off than any of the other other APIs at worse, but that
you can easily write more data to your dataset is quite nice.

-Jason

On Sun, Feb 12, 2017 at 1:23 PM, Serti Ayoub <[email protected]> wrote:

> Hello All,
>
> I'm new to HDF5 file format and I didn't see the difference between Table,
> DataSet and PacketTable? what's the difference between all those structure?
> And what every one stand for?
>
> Thank you in advance.
> Regards,
> Ayoub
>
>
>
>  Sent with Mailtrack
> <https://mailtrack.io/install?source=signature&lang=en&[email protected]&idSignature=22>
>
> _______________________________________________
> Hdf-forum is for HDF software users discussion.
> [email protected]
> http://lists.hdfgroup.org/mailman/listinfo/hdf-forum_lists.hdfgroup.org
> Twitter: https://twitter.com/hdf5
>
_______________________________________________
Hdf-forum is for HDF software users discussion.
[email protected]
http://lists.hdfgroup.org/mailman/listinfo/hdf-forum_lists.hdfgroup.org
Twitter: https://twitter.com/hdf5

Reply via email to