On Sat, Oct 29, 2016 at 11:44 PM, Stephan Hoyer firstname.lastname@example.org wrote:
I'm have more mixed fillings on testing for NaNs. NaNs propagate, so explicit testing is rarely needed. Also, in numerical computing we usually work with arrays of NaN, so operator.exists() and all this nice syntax would not be a substitute for numpy.isnan or pandas.isnull.
NaN's *usually* propagate. The NaN domain isn't actually closed under IEEE 754.
nan, inf = float('nan'), float('inf') import math nan**0
The last one isn't really mandated by IEEE 754, and is weird when you consider `min(nan, 1)`.