[Numpy-discussion] comparing arrays with NaN in them.

[Christopher Barker]

Timothy Hochberg wrote:
> in principle it's not safe to rely on NaNs being bitwise equal.

Thanks Tim, I always learn a lot on this list.

Anyway, I think my suggestion of "binary equal" wasn't really what I 
want. What I want is essentially a NaN-safe comparison, much like the 
NaN-safe functions like nanmax, nanmin, etc.

I guess what that would involve is looping through the arrays, checking 
for "==", then checking if both are NaN if it returns false. (or 
checking for the NaN's first).

I'm not sure NaNifying the other comparisons would make any sense: NaN > 
NaN (and all the other comparison's would have to return False anyway.

So, would a nanequal function be useful? would it be hard to write?

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov