economical min

[smichr]

Here's a function that will apply multiple keys economically to find the minimum of a list of elements. It might be useful to fight the canonical battles:

def lazymin(seq, keys):

    """Return minimum item using one or more keys that are provided in

    the list, applying keys only when necessary.

    Examples

    ========

    >>> from sympy.utilities.iterables import lazymin

    >>> from sympy import S

    >>> lazymin(range(20)+[S.Zero], keys=[int, hash])

    [0]

    >>> type(_[0])

    <class 'sympy.core.numbers.Zero'>

    """

    from collections import defaultdict

    if not seq:

        raise ValueError('empty sequence')

    while True:

        if len(seq) == 1:

            return seq[0]

        if not keys:

            raise ValueError("not enough keys to resolve min element")

        k = keys.pop() # error will be raised if we run out

        d=defaultdict(list)

        for s in seq:

            d[k(s)].append(s)

        seq = d[min(d)]

The first key is applied and if there is a tie for first place, the next key is applied only to those that tied, etc... until keys are exhausted or there is a winner.