On Sat, Mar 3, 2012 at 4:46 PM, Mark Wiebe <mwwi...@gmail.com> wrote: > On Sat, Mar 3, 2012 at 12:30 PM, Travis Oliphant <tra...@continuum.io> <snip> >> >> * the reduction operations need to default to "skipna" --- this is >> the most common use case which has been re-inforced again to me today by a >> new user to Python who is using masked arrays presently > > > This is a completely trivial change. I went with the default as I did > because it's what R, the primary inspiration for the NA design, does. We'll > have to be sure this is well-marked in the documentation about "NumPy NA for > R users". >
It may be trivial to change the code, but this isn't a trivial change. "Most common use case" is hard for me to swallow, since there are so many. Of the different statistical softwares I've used, none that I recall ignores missing data (silently) by default. This sounds dangerous to me. It's one thing to be convenient to work with missing data, but it's another to try to sweep the problem under the rug. I imagine the choice of the R developers was a thoughtful one. Perhaps something like np.seterr should also be implemented for missing data since there's probably no resolution to what's most sensible here. Skipper _______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion