On 12/6/06, Robert Kern <robert.kern@[...].com>  wrote:
>  Filip Wasilewski wrote:
> 
>  > Just like for other Python objects:
>  >
>  > if ifs3:
>  >    print "not empty"
> 
>  No, that doesn't work. numpy arrays do not have a truth value. They raise an
>  error when you try to use them in such a context.

Right!

I could swear I have checked this before posting. Evidently I got
bitten by this:

> >> bool(numpy.array([]))
False
> >> bool(numpy.array([1]))
True
> >> bool(numpy.array([0]))
False
> >> bool(numpy.array([1,1]))

Traceback (most recent call last):
  File "<pyshell#8> ", line 1, in -toplevel-
    bool(numpy.array([1,1]))
ValueError: The truth value of an array with more than one element is
ambiguous. Use a.any() or a.all()

So depending on the situation one can use len or size:
> >> len(numpy.array([[],[]]))
2
> >> numpy.array([[],[]]).size
0


And how to understand following?

> >> print numpy.array([1,1]) == [], numpy.array([1,1]) != []
False True
> >> print `numpy.array([1]) == []`, `numpy.array([1]) != []`
array([], dtype=bool) array([], dtype=bool)
> >> print bool(numpy.array([1]) == []), bool(numpy.array([1]) != [])
False False


cheers,
fw
_______________________________________________
Numpy-discussion mailing list
Numpy-discussion@[...].org
http://projects.scipy.org/mailman/listinfo/numpy-discussion