I was wondering if anyone could help explain this.

I answered this question successfully, but the comparison kept failing due to whitespace differences with the 'golden' result.

I checked this discussion and no-one seems to have had the same problem. However, to get my code to complete I had to use the following

import numpy

def nformat(x, s):
    return "{:{w}d}.".format(int(x), w=s)

a = numpy.array(input().split(), float)
for f in [numpy.floor, numpy.ceil, numpy.rint]:
    b = f(a)
    size = len(str(max(b)))-1
    numpy.set_printoptions(formatter={'all':lambda x: nformat(x, size)})        
    print(b)

There's probably many other ways to do it, and I won't claim this is the cleanest way - but essentially I had to muck with 'set_printoptions(formatter=...)' before each output was displayed to adjust the formatting width to accomodate the 'width' of the largest value.

Alternatively I supposed I could have written my own 'printN' function.

I was hoping someone might be able to explain why I had this problem and it seems no one else did (or didn't comment on it).

Thanks in advance,