[Pytables-users] direct numpy ndarray access

[filc e. <eka...@gm...>]

I have a numpy flavor Array in an HDF5 file.  I would like to
access an ndarray object representing the array and do
something like, x[1,2] = 3, and have that change reflected
in the file.  I want to use the ndarray methods to get numpy
to reinterpret the format of the data before I access it.
I am not sure how to do it.  I can use Array.read() to get
an ndarray object, but it appears that is a copy of the data,
and does not provide me direct access to bytes in the file.
I want to be able to access large (gigabytes) arrays of bytes,
so I do not want to read the whole thing, then write the
whole thing back.  I realize of course that the Array class
provides __getitem__ and __setitem__ style access, but I
need to tell the ndarray object the layout of the data.