I l@ve RuBoard Previous Section Next Section

17.3 Evaluating Predicate Tests Across Sequences

Credit: Jon Dyte

17.3.1 Problem

You need to know whether all items in a sequence satisfy a certain predicate or if only some of them do.

17.3.2 Solution

The simplest approach for either problem is to loop on the sequence and return a result as soon as it's known, just as the and and or Python operators short-circuit logical evaluation:

def every (pred, seq):
    """ true if pred(x) is true for all x in seq, else false """
    for x in seq:
        if not pred(x): return 0
    return 1

def any (pred, seq):
    """ false if pred(x) is false for all x in seq, else true """
    for x in seq:
        if pred(x): return 1
    return 0

17.3.3 Discussion

Often, it is useful to know whether all elements of a sequence meet certain criteria or if only some do. The two functions every and any do just that, with the simplest, and thus clearest, approach:

>>> every(lambda c: c > 5, (6, 7, 8, 9))
1
>>> every(lambda c: c > 5, (6, 4, 8, 9))
0
>>> any(lambda c: c > 5, (6, 7, 8, 9))
1
>>> any(lambda c: c < 5, (6, 7, 8, 9))
0

If you want to get fancy, here are two more techniques (perhaps with some performance gain, though you shouldn't take my word for it梐lways measure performance for the specific cases you need):

def every(pred, seq): return len(seq) == len(filter(pred, seq))
def any(pred, seq): return len(filter(pred, seq))

or:

import operator
def every(pred, seq):
    return reduce(operator.and_, map(pred, seq))
def any(pred, seq):
    return reduce(operator.or_, map(pred, seq))

Functional forms are elegant and often fast, as long as they do not involve a lambda.

    I l@ve RuBoard Previous Section Next Section