max() of big nested lists

Question

I encountered a quite odd problem while dealing with the max of a long list of lists of of pairs, such as

[
    [(0, 1), (1, 1), (2, 1), (3, 4), (4, 1), (5, 1), (6, 1),...,(141,3)],
    ..., 
    [(12, 1), (36, 1), (91, 1), (92, 1), (110, 1),..., (180, 1)]
]

I am trying to get the maximum of the first element of all the pairs. Pythonically, I was doing:

max([max(x) for x in list])[0]

which actually returns the correct number, IF the list is shorter than 281 lists. In fact, as soon as the list is longer than 280, I get this message

ValueError: max() arg is an empty sequence

So, for a long list

max([max(x) for x in list[0:280]])[0]

it's fine, while

max([max(x) for x in list[0:281]])[0]

breaks.

Am I doing something wrong here?

Answer 1

You have an empty list among your list of lists, at index 280. Slicing up to [:280] excludes it, and it is included with [:281] .

This is easily reproduced with a shorter sample:

>>> lsts = [
...     [(0, 1), (1, 1)],
...     [(2, 1), (3, 4)],
...     [(4, 1), (5, 1)],
...     [],  # empty at index 3
... ]
>>> max(max(x) for x in lsts)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 1, in <genexpr>
ValueError: max() arg is an empty sequence
>>> max(max(x) for x in lsts[:3])  # include everything before index 3
(5, 1)

You can avoid the issue altogether by chaining your lists together, here using chain.from_iterable() :

from itertools import chain

max(chain.from_iterable(list_of_lists))[0]

This treats all nested lists as one long list, an empty list somewhere in between simply doesn't contribute to that new sequence.

Answer 2

Why not just this?

max([max([t[0] for t in sl if t]) for sl in l if sl])

You can extract the first item from the beginning. Empty lists and tuples are ignored.

EDIT

max([max([t for t in sl if t]) for sl in l if sl])[0]

is more efficient.