numpy.linalg.norm gives weird result

Question

The following gives error:

print numpy.linalg.norm(2) # returns 2
print numpy.linalg.norm(2, np.inf) # returns error,
print numpy.linalg.norm(2, np.inf) # returns the same error:

ValueError: Improper number of dimensions to norm.

How can I use the above norm with non numpy array input?

Answer 1

As the doc string states:

In [165]: np.linalg.norm?
Definition: np.linalg.norm(x, ord=None, axis=None)   
...    
Parameters
----------
x : array_like
    Input array.  If `axis` is None, `x` must be 1-D or 2-D.

the first argument to norm should be an array_like object. Therefore use

In [167]: np.linalg.norm([2], np.inf)
Out[167]: 2