[Nipy-devel] good format for storing numeric data

Jonathan Taylor jonathan.taylor@stanford....
Mon Nov 16 14:17:28 CST 2009


hi satra,

>
> thanks. at present npz and npy probably make most sense. however, if,
> as you point earlier, hdf5 is the more general format (even though it
> might be overkill) and given that h5py supports numpy datatypes, do
> you think it makes sense to adopt hdf5 for binary storage within NiPy
> where standard imaging fileformats, like nifti, don't suffice?  (e.g.,
> i may want to store the output of an SVD operation on a numpy array).
>
>
to start off, i should point out that i don't know much about hdf5 or its
implementation in h5py.

if you do want to store
more complicated collections of arrays, then something like hdf5 does seem
(to me) more portable
than npy or npz. my comment about overkill referred to the fact that it
seems your requirement
that the format be well-supported in things like excel suggests that the
data structure you want to store is fairly simple and hdf5 seems much richer
than anything you would want to see in excel.

-- jonathan




-- 
Jonathan Taylor
Dept. of Statistics
Sequoia Hall, 137
390 Serra Mall
Stanford, CA 94305
Tel:   650.723.9230
Fax:   650.725.8977
Web: http://www-stat.stanford.edu/~jtaylo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.scipy.org/pipermail/nipy-devel/attachments/20091116/2b51163b/attachment.html 


More information about the Nipy-devel mailing list