On Fri, Apr 6, 2012 at 3:50 PM, Charles R Harris <charlesr.har...@gmail.com>wrote:
> > > On Fri, Apr 6, 2012 at 3:57 AM, Nathaniel Smith <n...@pobox.com> wrote: > >> On Fri, Apr 6, 2012 at 7:19 AM, Travis Oliphant <tra...@continuum.io> >> wrote: >> > That is an interesting point of view. I could see that point of >> view. >> > But, was this discussed as a bug prior to this change occurring? >> > >> > I just heard from a very heavy user of NumPy that they are nervous about >> > upgrading because of little changes like this one. I don't know if >> this >> > particular issue would affect them or not, but I will re-iterate my view >> > that we should be very careful of these kinds of changes. >> >> I agree -- these changes make me very nervous as well, especially >> since I haven't seen any short, simple description of what changed or >> what the rules actually are now (comparable to the old "scalars do not >> affect the type of arrays"). >> >> But, I also want to speak up in favor in one respect, since real world >> data points are always good. I had some code that did >> def do_something(a): >> a = np.asarray(a) >> a -= np.mean(a) >> ... >> If someone happens to pass in an integer array, then this is totally >> broken -- np.mean(a) may be non-integral, and in 1.6, numpy silently >> discards the fractional part and performs the subtraction anyway, >> e.g.: >> >> In [4]: a >> Out[4]: array([0, 1, 2, 3]) >> >> In [5]: a -= 1.5 >> >> In [6]: a >> Out[6]: array([-1, 0, 0, 1]) >> >> The bug was discovered when Skipper tried running my code against >> numpy master, and it errored out on the -=. So Mark's changes did >> catch one real bug that would have silently caused completely wrong >> numerical results! >> > As a second datapoint, it did catch real bugs in scikit-learn too. On the other hand, it required a workaround in ndimage. http://thread.gmane.org/gmane.comp.python.numeric.general/44206/focus=44208 > >> >> https://github.com/charlton/charlton/commit/d58c72529a5b33d06b49544bc3347c6480dc4512 >> >> Yes, these things are trade offs between correctness and convenience. I > don't mind new warnings/errors so much, they may break old code but they > don't lead to wrong results. It's the unexpected and unnoticed successes > that are scary. > We discussed reverting the unsafe casting behavior for 1.7 in the thread I linked to above. Do we still want to do this? As far as I can tell it didn't really cause problems so far. Ralf
_______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
