On 31 May 2011, at 17:45, Benjamin Root wrote:
>
> At this point, I wonder if it might be smarter to create
> a .atleast_Nd() function and use that everywhere it is needed.
> Having similar logic tailored for each loading function might be a
> little dangerous if bug fixes are made to one, but not the others.
Like a generalised version of .atleast_1d / .atleast_2d?
It would also have to include an .atmost_Nd functionality of some
sorts, to replace the .squeeze(), generally a good idea (e.g.
something like np.atleast_Nd(X, ndmin=0, ndmax=-1), where the default
is not to reduce the maximum number of dimensions...).
But for the io routines the situation is a bit more complex, since
different shapes are expected to be returned depending on the text
input (e.g. (1, N) for a single row vs. (N, 1) for a single column of
data).
Cheers,
Derek
_______________________________________________
NumPy-Discussion mailing list
NumPy-Discussion@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion
