python - Identity value for numpy ufunc reducers -

python - Identity value for numpy ufunc reducers -

i have numpy array want find maximum element of, call:

x = foo.max()

the problem foo empty array, , max function (understandably) throws:

valueerror: zero-size array reduction operation maximum has no identity

this leads question:

is there way supply identity value reduction operation? (ideally, 1 work arbitrary numpy ufuncs reduce methods.) or stuck doing:

if foo.size > 0: x = foo.max() else: x = 0 # or whatever identity value take

surely "pythonic" way go is:

def safe_max(ar,default=np.nan): try: homecoming np.max(ar) except valueerror: homecoming default x = safe_max(foo,0)

considering maximum of empty sequence not defined, seem logical approach.

python numpy