Python: How to loop a list of lists of varying depth?

Question

I have a data structure like this (the actual lists of lists are very long and of varying depth). I do know their depth beforehand.

 a=( [1,2], [2,3,[4,5]] )
 b=( [[1,2],[2,3]] )

A want to loop through each single list. How to best do this?

I don't want to end up doing something like this:

for l in a:
    if instance(l, list):
        for ll in l:
            if instance(ll, list): 
                ...

"I do know their depth beforehand." - is there a missing "not"? — Tim, Nov 11 '13 at 13:53
hint: the more appropriate name of this data structure is tree, rather than list — gefei, Nov 11 '13 at 13:54
@MartijnPieters, I want to pass the values to a `matplotlib` function — LarsVegas, Nov 11 '13 at 13:56
@MartijnPieters, I need to be sure the sequence doesn't contain another sequence before passing it to the function. Even more so, I need to pass each sequence to the function starting with the deepest level. — LarsVegas, Nov 11 '13 at 13:58
Why don't you edit your question with some more information based on the samples. What do you expect to happen for each given sample? — Martijn Pieters, Nov 11 '13 at 14:01

score 4 · Accepted Answer · answered Nov 11 '13 at 15:25

Since you don't defined the purpose, i'm coding a function that sum all elements:

def rec_sum(lst):
    if not lst:
        return 0
    el = lst.pop()
    if isinstance(el, list):
        return rec_sum(el) + rec_sum(lst)
    else:
        return el + rec_sum(lst)

Even if you know the depth beforehand, it's easier to solve using recursion.

Remember that Python limits 1000 stack frames stacked. So, if your list have more than 1000 items, you should get an Exception.

If you think you can have more than 1000 items, here it is a mixed solution, that uses recursion and for loops. It is limited into 1000 levels, instead of 1000 items:

def rec_for_sum(lst):
    if not lst:
        return 0
    count = 0
    for el in lst:
        if not isinstance(el, list):
            count += el
        else:
            count += rec_for_sum(el)
    return count

score 1 · Answer 2 · edited May 23 '17 at 10:31

1

You can check if an object is a list and just then go deeper:

Also see How to check if an object is a list or tuple (but not string)?

def myprint(list):
    for i in list:
        if isinstance(i, list):
            myprint(i)
        else:
            print i

edited May 23 '17 at 10:31

Community

1
1

answered Nov 11 '13 at 13:56

Florian Groetzner

512
3
12

This doesn't answer the question (insofar that the question can be understood at all); it should have been a comment instead. – Martijn Pieters Nov 11 '13 at 13:57
oh sry i clicked the wrong button first but now i edited it to a solution – Florian Groetzner Nov 11 '13 at 14:00

score 1 · Answer 3 · answered Nov 11 '13 at 14:51

You can use a combination of recursion and a generator:

def flatten_list(list_in):
    if isinstance(list_in,list):
        for l in list_in:
                for y in flatten_list(l):
                        yield y
    else:
        yield list_in


my_compound_list = [[1,2,3],[4,5,6],[7,8,9,[10,11,12,[13,14,15]]]]

print [f for f in flatten_list(my_compound_list)]

Jack_of_All_Trades · Answer 4 · 2013-11-11T14:22:02.447

0

SIMPLE FORM OF ITERATIVE SOLUTION (BASIS NOT COMPLETE):

   for item in your_list:
        if isinstance(item,list):
           ...........
        else:
        ...................

This is a way you should go. This should be enough to get you started.

RECURSIVE SOLUTION:

def list_flatten(my_list):
    for item in my_list:
        if(isinstance(item,list)):
            list_flatten(item)
        else:
            print(item)

This is a recursive solution for your problem. But beware of memory problem and recursion depth.

edited Nov 11 '13 at 14:22

answered Nov 11 '13 at 14:00

Jack_of_All_Trades

10,942
18
58
88

Ok, but if the the data tree has a depth of let's say 10? – LarsVegas Nov 11 '13 at 14:04

Python: How to loop a list of lists of varying depth?

4 Answers4

Linked