How to find positions of the list maximum?

Question

I have a list:

a = [32, 37, 28, 30, 37, 25, 27, 24, 35, 55, 23, 31, 55, 21, 40, 18, 50,
             35, 41, 49, 37, 19, 40, 41, 31]

max element is 55 (two elements on position 9 and 12)

I need to find on which position the maximum elements are situated. Please, help.

Answer 1

a.index(max(a))

will tell you the index of the first instance of the largest valued element of list a.

Answer 2

>>> m = max(a)
>>> [i for i, j in enumerate(a) if j == m]
[9, 12]

Answer 3

Here is the max value and the indexes it appears at:

>>> from collections import defaultdict
>>> d = defaultdict(list)
>>> a = [32, 37, 28, 30, 37, 25, 27, 24, 35, 55, 23, 31, 55, 21, 40, 18, 50, 35, 41, 49, 37, 19, 40, 41, 31]
>>> for i, x in enumerate(a):
...     d[x].append(i)
... 
>>> k = max(d.keys())
>>> print k, d[k]
55 [9, 12]

---

Later: for the satisfaction of @SilentGhost

>>> from itertools import takewhile
>>> import heapq
>>> 
>>> def popper(heap):
...     while heap:
...         yield heapq.heappop(heap)
... 
>>> a = [32, 37, 28, 30, 37, 25, 27, 24, 35, 55, 23, 31, 55, 21, 40, 18, 50, 35, 41, 49, 37, 19, 40, 41, 31]
>>> h = [(-x, i) for i, x in enumerate(a)]
>>> heapq.heapify(h)
>>> 
>>> largest = heapq.heappop(h)
>>> indexes = [largest[1]] + [x[1] for x in takewhile(lambda large: large[0] == largest[0], popper(h))]
>>> print -largest[0], indexes
55 [9, 12]

Answer 4

The chosen answer (and others) require at least two passes through the list. Here's a one pass solution.

Edited: To address the two deficiencies pointed out by @John Machin. For (2) I attempted to optimize the tests based on guesstimated probability of occurrence of each condition and inferences allowed from predecessors. It was a little tricky figuring out the proper initialization value formax_indices which worked for all possible cases, especially if the max happened to be the first value in the list.

a = [32, 37, 28, 30, 37, 25, 27, 24, 35, 55, 23, 31, 55, 21, 40, 18, 50,
             35, 41, 49, 37, 19, 40, 41, 31]

def maxelements(seq):
    ''' Return list of position(s) of largest element '''
    max_indices = []
    if len(seq):
        max_val = seq[0]
        for i,val in ((i,val) for i,val in enumerate(seq) if val >= max_val):
            if val == max_val:
                max_indices.append(i)
            else:
                max_val = val
                max_indices = [i]

    return max_indices

Answer 5

I can't reproduce the @SilentGhost-beating performance quoted by @martineau. Here's my effort with comparisons:

=== maxelements.py ===

a = [32, 37, 28, 30, 37, 25, 27, 24, 35, 55, 23, 31, 55, 21, 40, 18, 50,
             35, 41, 49, 37, 19, 40, 41, 31]
b = range(10000)
c = range(10000 - 1, -1, -1)
d = b + c

def maxelements_s(seq): # @SilentGhost
    ''' Return list of position(s) of largest element '''
    m = max(seq)
    return [i for i, j in enumerate(seq) if j == m]

def maxelements_m(seq): # @martineau
    ''' Return list of position(s) of largest element '''
    max_indices = []
    if len(seq):
        max_val = seq[0]
        for i, val in ((i, val) for i, val in enumerate(seq) if val >= max_val):
            if val == max_val:
                max_indices.append(i)
            else:
                max_val = val
                max_indices = [i]
    return max_indices

def maxelements_j(seq): # @John Machin
    ''' Return list of position(s) of largest element '''
    if not seq: return []
    max_val = seq[0] if seq[0] >= seq[-1] else seq[-1]
    max_indices = []
    for i, val in enumerate(seq):
        if val < max_val: continue
        if val == max_val:
            max_indices.append(i)
        else:
            max_val = val
            max_indices = [i]
    return max_indices

Results from a beat-up old laptop running Python 2.7 on Windows XP SP3:

>\python27\python -mtimeit -s"import maxelements as me" "me.maxelements_s(me.a)"
100000 loops, best of 3: 6.88 usec per loop

>\python27\python -mtimeit -s"import maxelements as me" "me.maxelements_m(me.a)"
100000 loops, best of 3: 11.1 usec per loop

>\python27\python -mtimeit -s"import maxelements as me" "me.maxelements_j(me.a)"
100000 loops, best of 3: 8.51 usec per loop

>\python27\python -mtimeit -s"import maxelements as me;a100=me.a*100" "me.maxelements_s(a100)"
1000 loops, best of 3: 535 usec per loop

>\python27\python -mtimeit -s"import maxelements as me;a100=me.a*100" "me.maxelements_m(a100)"
1000 loops, best of 3: 558 usec per loop

>\python27\python -mtimeit -s"import maxelements as me;a100=me.a*100" "me.maxelements_j(a100)"
1000 loops, best of 3: 489 usec per loop

Answer 6

import operator

def max_positions(iterable, key=None, reverse=False):
  if key is None:
    def key(x):
      return x
  if reverse:
    better = operator.lt
  else:
    better = operator.gt

  it = enumerate(iterable)
  for pos, item in it:
    break
  else:
    raise ValueError("max_positions: empty iterable")
    # note this is the same exception type raised by max([])
  cur_max = key(item)
  cur_pos = [pos]

  for pos, item in it:
    k = key(item)
    if better(k, cur_max):
      cur_max = k
      cur_pos = [pos]
    elif k == cur_max:
      cur_pos.append(pos)

  return cur_max, cur_pos

def min_positions(iterable, key=None, reverse=False):
  return max_positions(iterable, key, not reverse)

>>> L = range(10) * 2
>>> L
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> max_positions(L)
(9, [9, 19])
>>> min_positions(L)
(0, [0, 10])
>>> max_positions(L, key=lambda x: x // 2, reverse=True)
(0, [0, 1, 10, 11])