Splitting a list into N parts of approximately equal length

You can write it fairly simply as a list generator:

def split(a, n):
    k, m = divmod(len(a), n)
    return (a[i*k+min(i, m):(i+1)*k+min(i+1, m)] for i in range(n))

Example:

>>> list(split(range(11), 3))
[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10]]

As long as you don't want anything silly like continuous chunks:

>>> def chunkify(lst,n):
...     return [lst[i::n] for i in xrange(n)]
... 
>>> chunkify(range(13), 3)
[[0, 3, 6, 9, 12], [1, 4, 7, 10], [2, 5, 8, 11]]

This code is broken due to rounding errors. Do not use it!!!

assert len(chunkIt([1,2,3], 10)) == 10  # fails

Here's one that could work:

def chunkIt(seq, num):
    avg = len(seq) / float(num)
    out = []
    last = 0.0

    while last < len(seq):
        out.append(seq[int(last):int(last + avg)])
        last += avg

    return out

Testing:

>>> chunkIt(range(10), 3)
[[0, 1, 2], [3, 4, 5], [6, 7, 8, 9]]
>>> chunkIt(range(11), 3)
[[0, 1, 2], [3, 4, 5, 6], [7, 8, 9, 10]]
>>> chunkIt(range(12), 3)
[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10, 11]]

Changing the code to yield n chunks rather than chunks of n:

def chunks(l, n):
    """ Yield n successive chunks from l.
    """
    newn = int(len(l) / n)
    for i in xrange(0, n-1):
        yield l[i*newn:i*newn+newn]
    yield l[n*newn-newn:]

l = range(56)
three_chunks = chunks (l, 3)
print three_chunks.next()
print three_chunks.next()
print three_chunks.next()

which gives:

[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
[18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35]
[36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55]

This will assign the extra elements to the final group which is not perfect but well within your specification of "roughly N equal parts" :-) By that, I mean 56 elements would be better as (19,19,18) whereas this gives (18,18,20).

You can get the more balanced output with the following code:

#!/usr/bin/python
def chunks(l, n):
    """ Yield n successive chunks from l.
    """
    newn = int(1.0 * len(l) / n + 0.5)
    for i in xrange(0, n-1):
        yield l[i*newn:i*newn+newn]
    yield l[n*newn-newn:]

l = range(56)
three_chunks = chunks (l, 3)
print three_chunks.next()
print three_chunks.next()
print three_chunks.next()

which outputs:

[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]
[19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37]
[38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55]

This will do the split into equal parts by one single expression while keeping the order:

myList = list(range(18))  # given list
N = 5  # desired number of parts

[myList[(i*len(myList))//N:((i+1)*len(myList))//N] for i in range(N)]
# [[0, 1, 2], [3, 4, 5, 6], [7, 8, 9], [10, 11, 12, 13], [14, 15, 16, 17]]

The parts will differ in not more than one element. The split of 18 into 5 parts results in 3 + 4 + 3 + 4 + 4 = 18.

See more_itertools.divide:

n = 2

[list(x) for x in mit.divide(n, range(5, 11))]
# [[5, 6, 7], [8, 9, 10]]

[list(x) for x in mit.divide(n, range(5, 12))]
# [[5, 6, 7, 8], [9, 10, 11]]

Install via > pip install more_itertools.

Here is one that adds None to make the lists equal length

>>> from itertools import izip_longest
>>> def chunks(l, n):
    """ Yield n successive chunks from l. Pads extra spaces with None
    """
    return list(zip(*izip_longest(*[iter(l)]*n)))

>>> l=range(54)

>>> chunks(l,3)
[(0, 3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45, 48, 51), (1, 4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 34, 37, 40, 43, 46, 49, 52), (2, 5, 8, 11, 14, 17, 20, 23, 26, 29, 32, 35, 38, 41, 44, 47, 50, 53)]

>>> chunks(l,4)
[(0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52), (1, 5, 9, 13, 17, 21, 25, 29, 33, 37, 41, 45, 49, 53), (2, 6, 10, 14, 18, 22, 26, 30, 34, 38, 42, 46, 50, None), (3, 7, 11, 15, 19, 23, 27, 31, 35, 39, 43, 47, 51, None)]

>>> chunks(l,5)
[(0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50), (1, 6, 11, 16, 21, 26, 31, 36, 41, 46, 51), (2, 7, 12, 17, 22, 27, 32, 37, 42, 47, 52), (3, 8, 13, 18, 23, 28, 33, 38, 43, 48, 53), (4, 9, 14, 19, 24, 29, 34, 39, 44, 49, None)]

Have a look at numpy.split:

>>> a = numpy.array([1,2,3,4])
>>> numpy.split(a, 2)
[array([1, 2]), array([3, 4])]

Let's say you want to split a list [1, 2, 3, 4, 5, 6, 7, 8] into 3 element lists

like [[1,2,3], [4, 5, 6], [7, 8]], where if the last remaining elements left are less than 3, they are grouped together.

my_list = [1, 2, 3, 4, 5, 6, 7, 8]
my_list2 = [my_list[i:i+3] for i in range(0, len(my_list), 3)]
print(my_list2)

Output: [[1,2,3], [4, 5, 6], [7, 8]]

Where length of one part is 3. Replace 3 with your own chunk size.

My solution, easy to understand

def split_list(lst, n):
    splitted = []
    for i in reversed(range(1, n + 1)):
        split_point = len(lst)//i
        splitted.append(lst[:split_point])
        lst = lst[split_point:]
    return splitted

And shortest one-liner on this page(written by my girl)

def split(l, n):
    return [l[int(i*len(l)/n):int((i+1)*len(l)/n-1)] for i in range(n)]

Elegant:

[x.tolist() for x in np.array_split(range(10), 3)]

Another way would be something like this, the idea here is to use grouper, but get rid of None. In this case we'll have all 'small_parts' formed from elements at the first part of the list, and 'larger_parts' from the later part of the list. Length of 'larger parts' is len(small_parts) + 1. We need to consider x as two different sub-parts.

from itertools import izip_longest

import numpy as np

def grouper(n, iterable, fillvalue=None): # This is grouper from itertools
    "grouper(3, 'ABCDEFG', 'x') --> ABC DEF Gxx"
    args = [iter(iterable)] * n
    return izip_longest(fillvalue=fillvalue, *args)

def another_chunk(x,num):
    extra_ele = len(x)%num #gives number of parts that will have an extra element 
    small_part = int(np.floor(len(x)/num)) #gives number of elements in a small part

    new_x = list(grouper(small_part,x[:small_part*(num-extra_ele)]))
    new_x.extend(list(grouper(small_part+1,x[small_part*(num-extra_ele):])))

    return new_x

The way I have it set up returns a list of tuples:

>>> x = range(14)
>>> another_chunk(x,3)
[(0, 1, 2, 3), (4, 5, 6, 7, 8), (9, 10, 11, 12, 13)]
>>> another_chunk(x,4)
[(0, 1, 2), (3, 4, 5), (6, 7, 8, 9), (10, 11, 12, 13)]
>>> another_chunk(x,5)
[(0, 1), (2, 3, 4), (5, 6, 7), (8, 9, 10), (11, 12, 13)]
>>> 

#!/usr/bin/python


first_names = ['Steve', 'Jane', 'Sara', 'Mary','Jack','Bob', 'Bily', 'Boni', 'Chris','Sori', 'Will', 'Won','Li']

def chunks(l, n):
for i in range(0, len(l), n):
    # Create an index range for l of n items:
    yield l[i:i+n]

result = list(chunks(first_names, 5))
print result

Picked from this link, and this was what helped me. I had a pre-defined list.

def evenly(l, n):
    len_ = len(l)
    split_size = len_ // n
    split_size = n if not split_size else split_size
    offsets = [i for i in range(0, len_, split_size)]
    return [l[offset:offset + split_size] for offset in offsets]

Example:

l = [a for a in range(97)] should be consist of 10 parts, each have 9 elements except the last one.

Output:

[[0, 1, 2, 3, 4, 5, 6, 7, 8],
 [9, 10, 11, 12, 13, 14, 15, 16, 17],
 [18, 19, 20, 21, 22, 23, 24, 25, 26],
 [27, 28, 29, 30, 31, 32, 33, 34, 35],
 [36, 37, 38, 39, 40, 41, 42, 43, 44],
 [45, 46, 47, 48, 49, 50, 51, 52, 53],
 [54, 55, 56, 57, 58, 59, 60, 61, 62],
 [63, 64, 65, 66, 67, 68, 69, 70, 71],
 [72, 73, 74, 75, 76, 77, 78, 79, 80],
 [81, 82, 83, 84, 85, 86, 87, 88, 89],
 [90, 91, 92, 93, 94, 95, 96]]

1>

import numpy as np

data # your array

total_length = len(data)
separate = 10
sub_array_size = total_length // separate
safe_separate = sub_array_size * separate

splited_lists = np.split(np.array(data[:safe_separate]), separate)
splited_lists[separate - 1] = np.concatenate(splited_lists[separate - 1], 
np.array(data[safe_separate:total_length]))

splited_lists # your output

2>

splited_lists = np.array_split(np.array(data), separate)

def chunkify(target_list, chunk_size):
    return [target_list[i:i+chunk_size] for i in range(0, len(target_list), chunk_size)]

>>> l = [5432, 432, 67, "fdas", True, True, False, (4324,131), 876, "ofsa", 8, 909, b'765']
>>> print(chunkify(l, 3))
>>> [[5432, 432, 67], ['fdas', True, True], [False, (4324, 131), 876], ['ofsa', 8, 909], [b'765']]

say you want to split into 5 parts:

p1, p2, p3, p4, p5 = np.split(df, 5)

I've written code in this case myself:

def chunk_ports(port_start, port_end, portions):
    if port_end < port_start:
        return None

    total = port_end - port_start + 1

    fractions = int(math.floor(float(total) / portions))

    results = []

    # No enough to chuck.
    if fractions < 1:
        return None

    # Reverse, so any additional items would be in the first range.
    _e = port_end
    for i in range(portions, 0, -1):
        print "i", i

        if i == 1:
            _s = port_start
        else:
            _s = _e - fractions + 1

        results.append((_s, _e))

        _e = _s - 1

    results.reverse()

    return results

divide_ports(1, 10, 9) would return

[(1, 2), (3, 3), (4, 4), (5, 5), (6, 6), (7, 7), (8, 8), (9, 9), (10, 10)]

this code works for me (Python3-compatible):

def chunkify(tab, num):
    return [tab[i*num: i*num+num] for i in range(len(tab)//num+(1 if len(tab)%num else 0))]

example (for bytearray type, but it works for lists as well):

b = bytearray(b'\x01\x02\x03\x04\x05\x06\x07\x08')
>>> chunkify(b,3)
[bytearray(b'\x01\x02\x03'), bytearray(b'\x04\x05\x06'), bytearray(b'\x07\x08')]
>>> chunkify(b,4)
[bytearray(b'\x01\x02\x03\x04'), bytearray(b'\x05\x06\x07\x08')]

This one provides chunks of length <= n, >= 0

def

 chunkify(lst, n):
    num_chunks = int(math.ceil(len(lst) / float(n))) if n < len(lst) else 1
    return [lst[n*i:n*(i+1)] for i in range(num_chunks)]

for example

>>> chunkify(range(11), 3)
[[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10]]
>>> chunkify(range(11), 8)
[[0, 1, 2, 3, 4, 5, 6, 7], [8, 9, 10]]

I tried most part of solutions, but they didn't work for my case, so I make a new function that work for most of cases and for any type of array:

import math

def chunkIt(seq, num):
    seqLen = len(seq)
    total_chunks = math.ceil(seqLen / num)
    items_per_chunk = num
    out = []
    last = 0

    while last < seqLen:
        out.append(seq[last:(last + items_per_chunk)])
        last += items_per_chunk

    return out

def chunk_array(array : List, n: int) -> List[List]:
    chunk_size = len(array) // n 
    chunks = []
    i = 0
    while i < len(array):
        # if less than chunk_size left add the remainder to last element
        if len(array) - (i + chunk_size + 1) < 0:
            chunks[-1].append(*array[i:i + chunk_size])
            break
        else:
            chunks.append(array[i:i + chunk_size])
            i += chunk_size
    return chunks

here's my version (inspired from Max's)

n = len(lst)
# p is the number of parts to be divided
x = int(n/p)

i = 0
j = x
lstt = []
while (i< len(lst) or j <len(lst)):
    lstt.append(lst[i:j])
    i+=x
    j+=x
print(lstt)

This is the simplest answer if it is known that the list divides into equal parts.

If you don't mind that the order will be changed, I recommend you to use @job solution, otherwise, you can use this:

def chunkIt(seq, num):
    steps = int(len(seq) / float(num))
    out = []
    last = 0.0

    while last < len(seq):
        if len(seq) - (last + steps) < steps:
            until = len(seq)
            steps = len(seq) - last
        else:
            until = int(last + steps)
        out.append(seq[int(last): until])
        last += steps
return out

Another attempt at simple readable chunker that works.

def chunk(iterable, count): # returns a *generator* that divides `iterable` into `count` of contiguous chunks of similar size
    assert count >= 1
    return (iterable[int(_*len(iterable)/count+0.5):int((_+1)*len(iterable)/count+0.5)] for _ in range(count))

print("Chunk count:  ", len(list(         chunk(range(105),10))))
print("Chunks:       ",     list(         chunk(range(105),10)))
print("Chunks:       ",     list(map(list,chunk(range(105),10))))
print("Chunk lengths:",     list(map(len, chunk(range(105),10))))

print("Testing...")
for iterable_length in range(100):
    for chunk_count in range(1,100):
        chunks = list(chunk(range(iterable_length),chunk_count))
        assert chunk_count == len(chunks)
        assert iterable_length == sum(map(len,chunks))
        assert all(map(lambda _:abs(len(_)-iterable_length/chunk_count)<=1,chunks))
print("Okay")

Outputs:

Chunk count:   10
Chunks:        [range(0, 11), range(11, 21), range(21, 32), range(32, 42), range(42, 53), range(53, 63), range(63, 74), range(74, 84), range(84, 95), range(95, 105)]
Chunks:        [[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], [11, 12, 13, 14, 15, 16, 17, 18, 19, 20], [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31], [32, 33, 34, 35, 36, 37, 38, 39, 40, 41], [42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52], [53, 54, 55, 56, 57, 58, 59, 60, 61, 62], [63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73], [74, 75, 76, 77, 78, 79, 80, 81, 82, 83], [84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94], [95, 96, 97, 98, 99, 100, 101, 102, 103, 104]]
Chunk lengths: [11, 10, 11, 10, 11, 10, 11, 10, 11, 10]
Testing...
Okay