Related
Given a set
{0, 1, 2, 3}
How can I produce the subsets:
[set(),
{0},
{1},
{2},
{3},
{0, 1},
{0, 2},
{0, 3},
{1, 2},
{1, 3},
{2, 3},
{0, 1, 2},
{0, 1, 3},
{0, 2, 3},
{1, 2, 3},
{0, 1, 2, 3}]
The Python itertools page has exactly a powerset recipe for this:
from itertools import chain, combinations
def powerset(iterable):
"powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)"
s = list(iterable)
return chain.from_iterable(combinations(s, r) for r in range(len(s)+1))
Output:
>>> list(powerset("abcd"))
[(), ('a',), ('b',), ('c',), ('d',), ('a', 'b'), ('a', 'c'), ('a', 'd'), ('b', 'c'), ('b', 'd'), ('c', 'd'), ('a', 'b', 'c'), ('a', 'b', 'd'), ('a', 'c', 'd'), ('b', 'c', 'd'), ('a', 'b', 'c', 'd')]
If you don't like that empty tuple at the beginning, you can just change the range statement to range(1, len(s)+1) to avoid a 0-length combination.
Here is more code for a powerset. This is written from scratch:
>>> def powerset(s):
... x = len(s)
... for i in range(1 << x):
... print [s[j] for j in range(x) if (i & (1 << j))]
...
>>> powerset([4,5,6])
[]
[4]
[5]
[4, 5]
[6]
[4, 6]
[5, 6]
[4, 5, 6]
Mark Rushakoff's comment is applicable here: "If you don't like that empty tuple at the beginning, on."you can just change the range statement to range(1, len(s)+1) to avoid a 0-length combination", except in my case you change for i in range(1 << x) to for i in range(1, 1 << x).
Returning to this years later, I'd now write it like this:
def powerset(s):
x = len(s)
masks = [1 << i for i in range(x)]
for i in range(1 << x):
yield [ss for mask, ss in zip(masks, s) if i & mask]
And then the test code would look like this, say:
print(list(powerset([4, 5, 6])))
Using yield means that you do not need to calculate all results in a single piece of memory. Precalculating the masks outside the main loop is assumed to be a worthwhile optimization.
If you're looking for a quick answer, I just searched "python power set" on google and came up with this: Python Power Set Generator
Here's a copy-paste from the code in that page:
def powerset(seq):
"""
Returns all the subsets of this set. This is a generator.
"""
if len(seq) <= 1:
yield seq
yield []
else:
for item in powerset(seq[1:]):
yield [seq[0]]+item
yield item
This can be used like this:
l = [1, 2, 3, 4]
r = [x for x in powerset(l)]
Now r is a list of all the elements you wanted, and can be sorted and printed:
r.sort()
print r
[[], [1], [1, 2], [1, 2, 3], [1, 2, 3, 4], [1, 2, 4], [1, 3], [1, 3, 4], [1, 4], [2], [2, 3], [2, 3, 4], [2, 4], [3], [3, 4], [4]]
Use function powerset() from package more_itertools.
Yields all possible subsets of the iterable
>>> list(powerset([1, 2, 3]))
[(), (1,), (2,), (3,), (1, 2), (1, 3), (2, 3), (1, 2, 3)]
If you want sets, use:
list(map(set, powerset(iterable)))
I have found the following algorithm very clear and simple:
def get_powerset(some_list):
"""Returns all subsets of size 0 - len(some_list) for some_list"""
if len(some_list) == 0:
return [[]]
subsets = []
first_element = some_list[0]
remaining_list = some_list[1:]
# Strategy: get all the subsets of remaining_list. For each
# of those subsets, a full subset list will contain both
# the original subset as well as a version of the subset
# that contains first_element
for partial_subset in get_powerset(remaining_list):
subsets.append(partial_subset)
subsets.append(partial_subset[:] + [first_element])
return subsets
Another way one can generate the powerset is by generating all binary numbers that have n bits. As a power set the amount of number with n digits is 2 ^ n. The principle of this algorithm is that an element could be present or not in a subset as a binary digit could be one or zero but not both.
def power_set(items):
N = len(items)
# enumerate the 2 ** N possible combinations
for i in range(2 ** N):
combo = []
for j in range(N):
# test bit jth of integer i
if (i >> j) % 2 == 1:
combo.append(items[j])
yield combo
I found both algorithms when I was taking MITx: 6.00.2x Introduction to Computational Thinking and Data Science, and I consider it is one of the easiest algorithms to understand I have seen.
from functools import reduce
def powerset(lst):
return reduce(lambda result, x: result + [subset + [x] for subset in result],
lst, [[]])
There is a refinement of powerset:
def powerset(seq):
"""
Returns all the subsets of this set. This is a generator.
"""
if len(seq) <= 0:
yield []
else:
for item in powerset(seq[1:]):
yield [seq[0]]+item
yield item
TL;DR (go directly to Simplification)
I know I have previously added an answer, but I really like my new implementation. I am taking a set as input, but it actually could be any iterable, and I am returning a set of sets which is the power set of the input. I like this approach because it is more aligned with the mathematical definition of power set (set of all subsets).
def power_set(A):
"""A is an iterable (list, tuple, set, str, etc)
returns a set which is the power set of A."""
length = len(A)
l = [a for a in A]
ps = set()
for i in range(2 ** length):
selector = f'{i:0{length}b}'
subset = {l[j] for j, bit in enumerate(selector) if bit == '1'}
ps.add(frozenset(subset))
return ps
If you want exactly the output you posted in your answer use this:
>>> [set(s) for s in power_set({1, 2, 3, 4})]
[{3, 4},
{2},
{1, 4},
{2, 3, 4},
{2, 3},
{1, 2, 4},
{1, 2},
{1, 2, 3},
{3},
{2, 4},
{1},
{1, 2, 3, 4},
set(),
{1, 3},
{1, 3, 4},
{4}]
Explanation
It is known that the number of elements of the power set is 2 ** len(A), so that could clearly be seen in the for loop.
I need to convert the input (ideally a set) into a list because by a set is a data structure of unique unordered elements, and the order will be crucial to generate the subsets.
selector is key in this algorithm. Note that selector has the same length as the input set, and to make this possible it is using an f-string with padding. Basically, this allows me to select the elements that will be added to each subset during each iteration. Let's say the input set has 3 elements {0, 1, 2}, so selector will take values between 0 and 7 (inclusive), which in binary are:
000 # 0
001 # 1
010 # 2
011 # 3
100 # 4
101 # 5
110 # 6
111 # 7
So, each bit could serve as an indicator if an element of the original set should be added or not. Look at the binary numbers, and just think of each number as an element of the super set in which 1 means that an element at index j should be added, and 0 means that this element should not be added.
I am using a set comprehension to generate a subset at each iteration, and I convert this subset into a frozenset so I can add it to ps (power set). Otherwise, I won't be able to add it because a set in Python consists only of immutable objects.
Simplification
You can simplify the code using some python comprehensions, so you can get rid of those for loops. You can also use zip to avoid using j index and the code will end up as the following:
def power_set(A):
length = len(A)
return {
frozenset({e for e, b in zip(A, f'{i:{length}b}') if b == '1'})
for i in range(2 ** length)
}
That's it. What I like of this algorithm is that is clearer and more intuitive than others because it looks quite magical to rely on itertools even though it works as expected.
def get_power_set(s):
power_set=[[]]
for elem in s:
# iterate over the sub sets so far
for sub_set in power_set:
# add a new subset consisting of the subset at hand added elem
power_set=power_set+[list(sub_set)+[elem]]
return power_set
For example:
get_power_set([1,2,3])
yield
[[], [1], [2], [1, 2], [3], [1, 3], [2, 3], [1, 2, 3]]
This can be done very naturally with itertools.product:
import itertools
def powerset(l):
for sl in itertools.product(*[[[], [i]] for i in l]):
yield {j for i in sl for j in i}
I know this is too late
There are many other solutions already but still...
def power_set(lst):
pw_set = [[]]
for i in range(0,len(lst)):
for j in range(0,len(pw_set)):
ele = pw_set[j].copy()
ele = ele + [lst[i]]
pw_set = pw_set + [ele]
return pw_set
I just wanted to provide the most comprehensible solution, the anti code-golf version.
from itertools import combinations
l = ["x", "y", "z", ]
def powerset(items):
combo = []
for r in range(len(items) + 1):
#use a list to coerce a actual list from the combinations generator
combo.append(list(combinations(items,r)))
return combo
l_powerset = powerset(l)
for i, item in enumerate(l_powerset):
print "All sets of length ", i
print item
The results
All sets of length 0
[()]
All sets of length 1
[('x',), ('y',), ('z',)]
All sets of length 2
[('x', 'y'), ('x', 'z'), ('y', 'z')]
All sets of length 3
[('x', 'y', 'z')]
For more see the itertools docs, also the wikipedia entry on power sets
With empty set, which is part of all the subsets, you could use:
def subsets(iterable):
for n in range(len(iterable) + 1):
yield from combinations(iterable, n)
from itertools import combinations
def subsets(arr: set) -> list:
subsets = []
[subsets.extend(list(combinations(arr,n))) for n in range(len(arr))]
return subsets
a = {1,2,3}
print(subsets(a))
Output:
[(), (1,), (2,), (3,), (1, 2), (1, 3), (2, 3)]
For sorted subsets, we can do:
# sorted subsets
print(sorted(subsets(a)))
Output:
[(), (1,), (1, 2), (1, 3), (2,), (2, 3), (3,)]
Just a quick power set refresher !
Power set of a set X, is simply the set of all subsets of X including
the empty set
Example set X = (a,b,c)
Power Set = { { a , b , c } , { a , b } , { a , c } , { b , c } , { a } , { b } , { c } , { } }
Here is another way of finding power set:
def power_set(input):
# returns a list of all subsets of the list a
if (len(input) == 0):
return [[]]
else:
main_subset = [ ]
for small_subset in power_set(input[1:]):
main_subset += [small_subset]
main_subset += [[input[0]] + small_subset]
return main_subset
print(power_set([0,1,2,3]))
full credit to source
If you want any specific length of subsets you can do it like this:
from itertools import combinations
someSet = {0, 1, 2, 3}
([x for i in range(len(someSet)+1) for x in combinations(someSet,i)])
More generally for arbitary length subsets you can modify the range arugment. The output is
[(), (0,), (1,), (2,), (3,), (0, 1), (0, 2), (0, 3), (1, 2), (1, 3), (2, 3), (0, 1, 2), (0, 1, 3), (0, 2, 3), (1, 2, 3), (0, 1, 2, 3)]
You can do it like this:
def powerset(x):
m=[]
if not x:
m.append(x)
else:
A = x[0]
B = x[1:]
for z in powerset(B):
m.append(z)
r = [A] + z
m.append(r)
return m
print(powerset([1, 2, 3, 4]))
Output:
[[], [1], [2], [1, 2], [3], [1, 3], [2, 3], [1, 2, 3], [4], [1, 4], [2, 4], [1, 2, 4], [3, 4], [1, 3, 4], [2, 3, 4], [1, 2, 3, 4]]
A simple way would be to harness the internal representation of integers under 2's complement arithmetic.
Binary representation of integers is as {000, 001, 010, 011, 100, 101, 110, 111} for numbers ranging from 0 to 7. For an integer counter value, considering 1 as inclusion of corresponding element in collection and '0' as exclusion we can generate subsets based on the counting sequence. Numbers have to be generated from 0 to pow(2,n) -1 where n is the length of array i.e. number of bits in binary representation.
A simple Subset Generator Function based on it can be written as below. It basically relies
def subsets(array):
if not array:
return
else:
length = len(array)
for max_int in range(0x1 << length):
subset = []
for i in range(length):
if max_int & (0x1 << i):
subset.append(array[i])
yield subset
and then it can be used as
def get_subsets(array):
powerset = []
for i in subsets(array):
powerser.append(i)
return powerset
Testing
Adding following in local file
if __name__ == '__main__':
sample = ['b', 'd', 'f']
for i in range(len(sample)):
print "Subsets for " , sample[i:], " are ", get_subsets(sample[i:])
gives following output
Subsets for ['b', 'd', 'f'] are [[], ['b'], ['d'], ['b', 'd'], ['f'], ['b', 'f'], ['d', 'f'], ['b', 'd', 'f']]
Subsets for ['d', 'f'] are [[], ['d'], ['f'], ['d', 'f']]
Subsets for ['f'] are [[], ['f']]
Almost all of these answers use list rather than set, which felt like a bit of a cheat to me. So, out of curiosity I tried to do a simple version truly on set and summarize for other "new to Python" folks.
I found there's a couple oddities in dealing with Python's set implementation. The main surprise to me was handling empty sets. This is in contrast to Ruby's Set implementation, where I can simply do Set[Set[]] and get a Set containing one empty Set, so I found it initially a little confusing.
To review, in doing powerset with sets, I encountered two problems:
set() takes an iterable, so set(set()) will return set() because the empty set iterable is empty (duh I guess :))
to get a set of sets, set({set()}) and set.add(set) won't work because set() isn't hashable
To solve both issues, I made use of frozenset(), which means I don't quite get what I want (type is literally set), but makes use of the overall set interace.
def powerset(original_set):
# below gives us a set with one empty set in it
ps = set({frozenset()})
for member in original_set:
subset = set()
for m in ps:
# to be added into subset, needs to be
# frozenset.union(set) so it's hashable
subset.add(m.union(set([member]))
ps = ps.union(subset)
return ps
Below we get 2² (16) frozensets correctly as output:
In [1]: powerset(set([1,2,3,4]))
Out[2]:
{frozenset(),
frozenset({3, 4}),
frozenset({2}),
frozenset({1, 4}),
frozenset({3}),
frozenset({2, 3}),
frozenset({2, 3, 4}),
frozenset({1, 2}),
frozenset({2, 4}),
frozenset({1}),
frozenset({1, 2, 4}),
frozenset({1, 3}),
frozenset({1, 2, 3}),
frozenset({4}),
frozenset({1, 3, 4}),
frozenset({1, 2, 3, 4})}
As there's no way to have a set of sets in Python, if you want to turn these frozensets into sets, you'll have to map them back into a list (list(map(set, powerset(set([1,2,3,4]))))
) or modify the above.
Perhaps the question is getting old, but I hope my code will help someone.
def powSet(set):
if len(set) == 0:
return [[]]
return addtoAll(set[0],powSet(set[1:])) + powSet(set[1:])
def addtoAll(e, set):
for c in set:
c.append(e)
return set
Getting all the subsets with recursion. Crazy-ass one-liner
from typing import List
def subsets(xs: list) -> List[list]:
return subsets(xs[1:]) + [x + [xs[0]] for x in subsets(xs[1:])] if xs else [[]]
Based on a Haskell solution
subsets :: [a] -> [[a]]
subsets [] = [[]]
subsets (x:xs) = map (x:) (subsets xs) ++ subsets xs
def findsubsets(s, n):
return list(itertools.combinations(s, n))
def allsubsets(s) :
a = []
for x in range(1,len(s)+1):
a.append(map(set,findsubsets(s,x)))
return a
def powerSet(s):
sets = [[]]
for i in s:
newsets = []
for k in sets:
newsets.append(k+[i])
sets += newsets
return sets
Code first, for those who want a simple answer.
I have a good explanation here https://leetcode.com/problems/subsets/solutions/3138042/simple-python-solution/
But the short answer is that you start with the set of the empty set, i.e. "sets = [[]]". I recommend to put a print a statement under "for i in s" i.e. "print(sets)" and see that it doubles for each element i
This is wild because none of these answers actually provide the return of an actual Python set. Here is a messy implementation that will give a powerset that actually is a Python set.
test_set = set(['yo', 'whatup', 'money'])
def powerset( base_set ):
""" modified from pydoc's itertools recipe shown above"""
from itertools import chain, combinations
base_list = list( base_set )
combo_list = [ combinations(base_list, r) for r in range(len(base_set)+1) ]
powerset = set([])
for ll in combo_list:
list_of_frozensets = list( map( frozenset, map( list, ll ) ) )
set_of_frozensets = set( list_of_frozensets )
powerset = powerset.union( set_of_frozensets )
return powerset
print powerset( test_set )
# >>> set([ frozenset(['money','whatup']), frozenset(['money','whatup','yo']),
# frozenset(['whatup']), frozenset(['whatup','yo']), frozenset(['yo']),
# frozenset(['money','yo']), frozenset(['money']), frozenset([]) ])
I'd love to see a better implementation, though.
Here is my quick implementation utilizing combinations but using only built-ins.
def powerSet(array):
length = str(len(array))
formatter = '{:0' + length + 'b}'
combinations = []
for i in xrange(2**int(length)):
combinations.append(formatter.format(i))
sets = set()
currentSet = []
for combo in combinations:
for i,val in enumerate(combo):
if val=='1':
currentSet.append(array[i])
sets.add(tuple(sorted(currentSet)))
currentSet = []
return sets
All subsets in range n as set:
n = int(input())
l = [i for i in range (1, n + 1)]
for number in range(2 ** n) :
binary = bin(number)[: 1 : -1]
subset = [l[i] for i in range(len(binary)) if binary[i] == "1"]
print(set(sorted(subset)) if number > 0 else "{}")
import math
def printPowerSet(set,set_size):
pow_set_size =int(math.pow(2, set_size))
for counter in range(pow_set_size):
for j in range(set_size):
if((counter & (1 << j)) > 0):
print(set[j], end = "")
print("")
set = ['a', 'b', 'c']
printPowerSet(set,3)
A variation of the question, is an exercise I see on the book "Discovering Computer Science: Interdisciplinary Problems, Principles, and Python Programming. 2015 edition". In that exercise 10.2.11, the input is just an integer number, and the output should be the power sets. Here is my recursive solution (not using anything else but basic python3 )
def powerSetR(n):
assert n >= 0
if n == 0:
return [[]]
else:
input_set = list(range(1, n+1)) # [1,2,...n]
main_subset = [ ]
for small_subset in powerSetR(n-1):
main_subset += [small_subset]
main_subset += [ [input_set[-1]] + small_subset]
return main_subset
superset = powerSetR(4)
print(superset)
print("Number of sublists:", len(superset))
And the output is
[[], [4], [3], [4, 3], [2], [4, 2], [3, 2], [4, 3, 2], [1], [4, 1], [3, 1], [4, 3, 1], [2, 1], [4, 2, 1], [3, 2, 1], [4, 3, 2, 1]]
Number of sublists: 16
I hadn't come across the more_itertools.powerset function and would recommend using that. I also recommend not using the default ordering of the output from itertools.combinations, often instead you want to minimise the distance between the positions and sort the subsets of items with shorter distance between them above/before the items with larger distance between them.
The itertools recipes page shows it uses chain.from_iterable
Note that the r here matches the standard notation for the lower part of a binomial coefficient, the s is usually referred to as n in mathematics texts and on calculators (“n Choose r”)
def powerset(iterable):
"powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)"
s = list(iterable)
return chain.from_iterable(combinations(s, r) for r in range(len(s)+1))
The other examples here give the powerset of [1,2,3,4] in such a way that the 2-tuples are listed in "lexicographic" order (when we print the numbers as integers). If I write the distance between the numbers alongside it (i.e. the difference), it shows my point:
12 ⇒ 1
13 ⇒ 2
14 ⇒ 3
23 ⇒ 1
24 ⇒ 2
34 ⇒ 1
The correct order for subsets should be the order which 'exhausts' the minimal distance first, like so:
12 ⇒ 1
23 ⇒ 1
34 ⇒ 1
13 ⇒ 2
24 ⇒ 2
14 ⇒ 3
Using numbers here makes this ordering look 'wrong', but consider for example the letters ["a","b","c","d"] it is clearer why this might be useful to obtain the powerset in this order:
ab ⇒ 1
bc ⇒ 1
cd ⇒ 1
ac ⇒ 2
bd ⇒ 2
ad ⇒ 3
This effect is more pronounced with more items, and for my purposes it makes the difference between being able to describe the ranges of the indexes of the powerset meaningfully.
(There is a lot written on Gray codes etc. for the output order of algorithms in combinatorics, I don't see it as a side issue).
I actually just wrote a fairly involved program which used this fast integer partition code to output the values in the proper order, but then I discovered more_itertools.powerset and for most uses it's probably fine to just use that function like so:
from more_itertools import powerset
from numpy import ediff1d
def ps_sorter(tup):
l = len(tup)
d = ediff1d(tup).tolist()
return l, d
ps = powerset([1,2,3,4])
ps = sorted(ps, key=ps_sorter)
for x in ps:
print(x)
⇣
()
(1,)
(2,)
(3,)
(4,)
(1, 2)
(2, 3)
(3, 4)
(1, 3)
(2, 4)
(1, 4)
(1, 2, 3)
(2, 3, 4)
(1, 2, 4)
(1, 3, 4)
(1, 2, 3, 4)
I wrote some more involved code which will print the powerset nicely (see the repo for pretty printing functions I've not included here: print_partitions, print_partitions_by_length, and pprint_tuple).
Repo: ordered-powerset, specifically pset_partitions.py
This is all pretty simple, but still might be useful if you want some code that'll let you get straight to accessing the different levels of the powerset:
from itertools import permutations as permute
from numpy import cumsum
# http://jeromekelleher.net/generating-integer-partitions.html
# via
# https://stackoverflow.com/questions/10035752/elegant-python-code-for-integer-partitioning#comment25080713_10036764
def asc_int_partitions(n):
a = [0 for i in range(n + 1)]
k = 1
y = n - 1
while k != 0:
x = a[k - 1] + 1
k -= 1
while 2 * x <= y:
a[k] = x
y -= x
k += 1
l = k + 1
while x <= y:
a[k] = x
a[l] = y
yield tuple(a[:k + 2])
x += 1
y -= 1
a[k] = x + y
y = x + y - 1
yield tuple(a[:k + 1])
# https://stackoverflow.com/a/6285330/2668831
def uniquely_permute(iterable, enforce_sort=False, r=None):
previous = tuple()
if enforce_sort: # potential waste of effort (default: False)
iterable = sorted(iterable)
for p in permute(iterable, r):
if p > previous:
previous = p
yield p
def sum_min(p):
return sum(p), min(p)
def partitions_by_length(max_n, sorting=True, permuting=False):
partition_dict = {0: ()}
for n in range(1,max_n+1):
partition_dict.setdefault(n, [])
partitions = list(asc_int_partitions(n))
for p in partitions:
if permuting:
perms = uniquely_permute(p)
for perm in perms:
partition_dict.get(len(p)).append(perm)
else:
partition_dict.get(len(p)).append(p)
if not sorting:
return partition_dict
for k in partition_dict:
partition_dict.update({k: sorted(partition_dict.get(k), key=sum_min)})
return partition_dict
def print_partitions_by_length(max_n, sorting=True, permuting=True):
partition_dict = partitions_by_length(max_n, sorting=sorting, permuting=permuting)
for k in partition_dict:
if k == 0:
print(tuple(partition_dict.get(k)), end="")
for p in partition_dict.get(k):
print(pprint_tuple(p), end=" ")
print()
return
def generate_powerset(items, subset_handler=tuple, verbose=False):
"""
Generate the powerset of an iterable `items`.
Handling of the elements of the iterable is by whichever function is passed as
`subset_handler`, which must be able to handle the `None` value for the
empty set. The function `string_handler` will join the elements of the subset
with the empty string (useful when `items` is an iterable of `str` variables).
"""
ps = {0: [subset_handler()]}
n = len(items)
p_dict = partitions_by_length(n-1, sorting=True, permuting=True)
for p_len, parts in p_dict.items():
ps.setdefault(p_len, [])
if p_len == 0:
# singletons
for offset in range(n):
subset = subset_handler([items[offset]])
if verbose:
if offset > 0:
print(end=" ")
if offset == n - 1:
print(subset, end="\n")
else:
print(subset, end=",")
ps.get(p_len).append(subset)
for pcount, partition in enumerate(parts):
distance = sum(partition)
indices = (cumsum(partition)).tolist()
for offset in range(n - distance):
subset = subset_handler([items[offset]] + [items[offset:][i] for i in indices])
if verbose:
if offset > 0:
print(end=" ")
if offset == n - distance - 1:
print(subset, end="\n")
else:
print(subset, end=",")
ps.get(p_len).append(subset)
if verbose and p_len < n-1:
print()
return ps
As an example, I wrote a CLI demo program which takes a string as a command line argument:
python string_powerset.py abcdef
⇣
a, b, c, d, e, f
ab, bc, cd, de, ef
ac, bd, ce, df
ad, be, cf
ae, bf
af
abc, bcd, cde, def
abd, bce, cdf
acd, bde, cef
abe, bcf
ade, bef
ace, bdf
abf
aef
acf
adf
abcd, bcde, cdef
abce, bcdf
abde, bcef
acde, bdef
abcf
abef
adef
abdf
acdf
acef
abcde, bcdef
abcdf
abcef
abdef
acdef
abcdef
Here it is my solutions, it is similar (conceptually) with the solution of lmiguelvargasf.
Let me say that
-[math item] by defintion the powerset do contain the empty set
-[personal taste] and also that I don't like using frozenset.
So the input is a list and the output will be a list of lists. The function could close earlier, but I like the element of the power set to be order lexicographically, that essentially means nicely.
def power_set(L):
"""
L is a list.
The function returns the power set, but as a list of lists.
"""
cardinality=len(L)
n=2 ** cardinality
powerset = []
for i in range(n):
a=bin(i)[2:]
subset=[]
for j in range(len(a)):
if a[-j-1]=='1':
subset.append(L[j])
powerset.append(subset)
#the function could stop here closing with
#return powerset
powerset_orderred=[]
for k in range(cardinality+1):
for w in powerset:
if len(w)==k:
powerset_orderred.append(w)
return powerset_orderred
I have a simple dictionary of the form D = {2: 4, 3: 2, 5: 1, 7: 1}. There are 4 keys which are the primes and 4 values which represent the maximum power. Based on that I want to generate all combinations of the form :
p1^range1 * p2^range2 * p3^range3 * p4^range4
where range1 to range4 are the ranges corresponding to the values of the dictionary of each prime. For p1 = 2 the range will be range(0, 4+1)
The most simplistic way would be a nested for loop with four levels :
for i1 in range(0, 4+1):
for i2 in range(0, 2+1):
for i3 in range(0, 1+1):
for i4 in range(0, 1+1):
n = 2**i1 * 3**i2 * 5**i3 * 7**i4
which is what I intend and it is correct but this is a non-elegant and mechanical way.
Another way is to use the itertools.product which works if all the ranges are the same. Basically it is as follows :
for x in itertools.product(range(0, 4+1), repeat=4):
print(x, end=' ')
However this will generate all the possible combinations including (2, 2, 3, 4),
(4, 4, 2, 3) or (2, 3, 4, 4) which are not valid in my case as the power of 5 (which is the third element in the tuple) will never be 3 and the power of 7 will never be 4. I want to limit somehow if possible the range of elements . So in this case the maximum tuple will be (4, 2, 1, 1) corresponding to the powers (values in dictionary D). So basically I need ONLY the range of tuples from (0,0,0,0) to (4,2,1,1).
Is there a way to achieve this without using for loops and do it in a similar approach as with itertools.product?
Many thanks in advance.
Actually,
As miradulo wrote the solution is very simple :
either use :
for x in itertools.product(*map(range, (5, 3, 2, 2))) :
print(x, end=' ')
or :
for x in itertools.product(range(5), range(3), range(2), range(2)) :
print(x, end=' ')
Both ways solve the problem.
While itertools is the way to go, this question is an excellent exercise for practicing generator functions:
def multirange(d):
if len(d) == 1:
for i in range(d[0]):
yield [i]
elif len(d) > 1:
for i in range(d[-1]):
for a in multirange(d[:-1]):
yield a + [i]
def multirange_b(d):
l = len(d)
products = [1]
for k in d:
products.append(products[-1]*k)
n = products[-1]
for i in range(n):
yield [(i%products[j+1])//products[j] for j in range(l)]
for l in multirange([4,3,2]):
print(l)
for l in multirange_b([4,3,2]):
print(l)
Given a set
{0, 1, 2, 3}
How can I produce the subsets:
[set(),
{0},
{1},
{2},
{3},
{0, 1},
{0, 2},
{0, 3},
{1, 2},
{1, 3},
{2, 3},
{0, 1, 2},
{0, 1, 3},
{0, 2, 3},
{1, 2, 3},
{0, 1, 2, 3}]
The Python itertools page has exactly a powerset recipe for this:
from itertools import chain, combinations
def powerset(iterable):
"powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)"
s = list(iterable)
return chain.from_iterable(combinations(s, r) for r in range(len(s)+1))
Output:
>>> list(powerset("abcd"))
[(), ('a',), ('b',), ('c',), ('d',), ('a', 'b'), ('a', 'c'), ('a', 'd'), ('b', 'c'), ('b', 'd'), ('c', 'd'), ('a', 'b', 'c'), ('a', 'b', 'd'), ('a', 'c', 'd'), ('b', 'c', 'd'), ('a', 'b', 'c', 'd')]
If you don't like that empty tuple at the beginning, you can just change the range statement to range(1, len(s)+1) to avoid a 0-length combination.
Here is more code for a powerset. This is written from scratch:
>>> def powerset(s):
... x = len(s)
... for i in range(1 << x):
... print [s[j] for j in range(x) if (i & (1 << j))]
...
>>> powerset([4,5,6])
[]
[4]
[5]
[4, 5]
[6]
[4, 6]
[5, 6]
[4, 5, 6]
Mark Rushakoff's comment is applicable here: "If you don't like that empty tuple at the beginning, on."you can just change the range statement to range(1, len(s)+1) to avoid a 0-length combination", except in my case you change for i in range(1 << x) to for i in range(1, 1 << x).
Returning to this years later, I'd now write it like this:
def powerset(s):
x = len(s)
masks = [1 << i for i in range(x)]
for i in range(1 << x):
yield [ss for mask, ss in zip(masks, s) if i & mask]
And then the test code would look like this, say:
print(list(powerset([4, 5, 6])))
Using yield means that you do not need to calculate all results in a single piece of memory. Precalculating the masks outside the main loop is assumed to be a worthwhile optimization.
If you're looking for a quick answer, I just searched "python power set" on google and came up with this: Python Power Set Generator
Here's a copy-paste from the code in that page:
def powerset(seq):
"""
Returns all the subsets of this set. This is a generator.
"""
if len(seq) <= 1:
yield seq
yield []
else:
for item in powerset(seq[1:]):
yield [seq[0]]+item
yield item
This can be used like this:
l = [1, 2, 3, 4]
r = [x for x in powerset(l)]
Now r is a list of all the elements you wanted, and can be sorted and printed:
r.sort()
print r
[[], [1], [1, 2], [1, 2, 3], [1, 2, 3, 4], [1, 2, 4], [1, 3], [1, 3, 4], [1, 4], [2], [2, 3], [2, 3, 4], [2, 4], [3], [3, 4], [4]]
Use function powerset() from package more_itertools.
Yields all possible subsets of the iterable
>>> list(powerset([1, 2, 3]))
[(), (1,), (2,), (3,), (1, 2), (1, 3), (2, 3), (1, 2, 3)]
If you want sets, use:
list(map(set, powerset(iterable)))
I have found the following algorithm very clear and simple:
def get_powerset(some_list):
"""Returns all subsets of size 0 - len(some_list) for some_list"""
if len(some_list) == 0:
return [[]]
subsets = []
first_element = some_list[0]
remaining_list = some_list[1:]
# Strategy: get all the subsets of remaining_list. For each
# of those subsets, a full subset list will contain both
# the original subset as well as a version of the subset
# that contains first_element
for partial_subset in get_powerset(remaining_list):
subsets.append(partial_subset)
subsets.append(partial_subset[:] + [first_element])
return subsets
Another way one can generate the powerset is by generating all binary numbers that have n bits. As a power set the amount of number with n digits is 2 ^ n. The principle of this algorithm is that an element could be present or not in a subset as a binary digit could be one or zero but not both.
def power_set(items):
N = len(items)
# enumerate the 2 ** N possible combinations
for i in range(2 ** N):
combo = []
for j in range(N):
# test bit jth of integer i
if (i >> j) % 2 == 1:
combo.append(items[j])
yield combo
I found both algorithms when I was taking MITx: 6.00.2x Introduction to Computational Thinking and Data Science, and I consider it is one of the easiest algorithms to understand I have seen.
from functools import reduce
def powerset(lst):
return reduce(lambda result, x: result + [subset + [x] for subset in result],
lst, [[]])
There is a refinement of powerset:
def powerset(seq):
"""
Returns all the subsets of this set. This is a generator.
"""
if len(seq) <= 0:
yield []
else:
for item in powerset(seq[1:]):
yield [seq[0]]+item
yield item
TL;DR (go directly to Simplification)
I know I have previously added an answer, but I really like my new implementation. I am taking a set as input, but it actually could be any iterable, and I am returning a set of sets which is the power set of the input. I like this approach because it is more aligned with the mathematical definition of power set (set of all subsets).
def power_set(A):
"""A is an iterable (list, tuple, set, str, etc)
returns a set which is the power set of A."""
length = len(A)
l = [a for a in A]
ps = set()
for i in range(2 ** length):
selector = f'{i:0{length}b}'
subset = {l[j] for j, bit in enumerate(selector) if bit == '1'}
ps.add(frozenset(subset))
return ps
If you want exactly the output you posted in your answer use this:
>>> [set(s) for s in power_set({1, 2, 3, 4})]
[{3, 4},
{2},
{1, 4},
{2, 3, 4},
{2, 3},
{1, 2, 4},
{1, 2},
{1, 2, 3},
{3},
{2, 4},
{1},
{1, 2, 3, 4},
set(),
{1, 3},
{1, 3, 4},
{4}]
Explanation
It is known that the number of elements of the power set is 2 ** len(A), so that could clearly be seen in the for loop.
I need to convert the input (ideally a set) into a list because by a set is a data structure of unique unordered elements, and the order will be crucial to generate the subsets.
selector is key in this algorithm. Note that selector has the same length as the input set, and to make this possible it is using an f-string with padding. Basically, this allows me to select the elements that will be added to each subset during each iteration. Let's say the input set has 3 elements {0, 1, 2}, so selector will take values between 0 and 7 (inclusive), which in binary are:
000 # 0
001 # 1
010 # 2
011 # 3
100 # 4
101 # 5
110 # 6
111 # 7
So, each bit could serve as an indicator if an element of the original set should be added or not. Look at the binary numbers, and just think of each number as an element of the super set in which 1 means that an element at index j should be added, and 0 means that this element should not be added.
I am using a set comprehension to generate a subset at each iteration, and I convert this subset into a frozenset so I can add it to ps (power set). Otherwise, I won't be able to add it because a set in Python consists only of immutable objects.
Simplification
You can simplify the code using some python comprehensions, so you can get rid of those for loops. You can also use zip to avoid using j index and the code will end up as the following:
def power_set(A):
length = len(A)
return {
frozenset({e for e, b in zip(A, f'{i:{length}b}') if b == '1'})
for i in range(2 ** length)
}
That's it. What I like of this algorithm is that is clearer and more intuitive than others because it looks quite magical to rely on itertools even though it works as expected.
def get_power_set(s):
power_set=[[]]
for elem in s:
# iterate over the sub sets so far
for sub_set in power_set:
# add a new subset consisting of the subset at hand added elem to it
# effectively doubling the sets resulting in the 2^n sets in the powerset of s.
power_set=power_set+[list(sub_set)+[elem]]
return power_set
For example:
get_power_set([1,2,3])
yield
[[], [1], [2], [1, 2], [3], [1, 3], [2, 3], [1, 2, 3]]
This can be done very naturally with itertools.product:
import itertools
def powerset(l):
for sl in itertools.product(*[[[], [i]] for i in l]):
yield {j for i in sl for j in i}
I know this is too late
There are many other solutions already but still...
def power_set(lst):
pw_set = [[]]
for i in range(0,len(lst)):
for j in range(0,len(pw_set)):
ele = pw_set[j].copy()
ele = ele + [lst[i]]
pw_set = pw_set + [ele]
return pw_set
I just wanted to provide the most comprehensible solution, the anti code-golf version.
from itertools import combinations
l = ["x", "y", "z", ]
def powerset(items):
combo = []
for r in range(len(items) + 1):
#use a list to coerce a actual list from the combinations generator
combo.append(list(combinations(items,r)))
return combo
l_powerset = powerset(l)
for i, item in enumerate(l_powerset):
print "All sets of length ", i
print item
The results
All sets of length 0
[()]
All sets of length 1
[('x',), ('y',), ('z',)]
All sets of length 2
[('x', 'y'), ('x', 'z'), ('y', 'z')]
All sets of length 3
[('x', 'y', 'z')]
For more see the itertools docs, also the wikipedia entry on power sets
With empty set, which is part of all the subsets, you could use:
def subsets(iterable):
for n in range(len(iterable) + 1):
yield from combinations(iterable, n)
from itertools import combinations
def subsets(arr: set) -> list:
subsets = []
[subsets.extend(list(combinations(arr,n))) for n in range(len(arr))]
return subsets
a = {1,2,3}
print(subsets(a))
Output:
[(), (1,), (2,), (3,), (1, 2), (1, 3), (2, 3)]
For sorted subsets, we can do:
# sorted subsets
print(sorted(subsets(a)))
Output:
[(), (1,), (1, 2), (1, 3), (2,), (2, 3), (3,)]
Just a quick power set refresher !
Power set of a set X, is simply the set of all subsets of X including
the empty set
Example set X = (a,b,c)
Power Set = { { a , b , c } , { a , b } , { a , c } , { b , c } , { a } , { b } , { c } , { } }
Here is another way of finding power set:
def power_set(input):
# returns a list of all subsets of the list a
if (len(input) == 0):
return [[]]
else:
main_subset = [ ]
for small_subset in power_set(input[1:]):
main_subset += [small_subset]
main_subset += [[input[0]] + small_subset]
return main_subset
print(power_set([0,1,2,3]))
full credit to source
If you want any specific length of subsets you can do it like this:
from itertools import combinations
someSet = {0, 1, 2, 3}
([x for i in range(len(someSet)+1) for x in combinations(someSet,i)])
More generally for arbitary length subsets you can modify the range arugment. The output is
[(), (0,), (1,), (2,), (3,), (0, 1), (0, 2), (0, 3), (1, 2), (1, 3), (2, 3), (0, 1, 2), (0, 1, 3), (0, 2, 3), (1, 2, 3), (0, 1, 2, 3)]
You can do it like this:
def powerset(x):
m=[]
if not x:
m.append(x)
else:
A = x[0]
B = x[1:]
for z in powerset(B):
m.append(z)
r = [A] + z
m.append(r)
return m
print(powerset([1, 2, 3, 4]))
Output:
[[], [1], [2], [1, 2], [3], [1, 3], [2, 3], [1, 2, 3], [4], [1, 4], [2, 4], [1, 2, 4], [3, 4], [1, 3, 4], [2, 3, 4], [1, 2, 3, 4]]
A simple way would be to harness the internal representation of integers under 2's complement arithmetic.
Binary representation of integers is as {000, 001, 010, 011, 100, 101, 110, 111} for numbers ranging from 0 to 7. For an integer counter value, considering 1 as inclusion of corresponding element in collection and '0' as exclusion we can generate subsets based on the counting sequence. Numbers have to be generated from 0 to pow(2,n) -1 where n is the length of array i.e. number of bits in binary representation.
A simple Subset Generator Function based on it can be written as below. It basically relies
def subsets(array):
if not array:
return
else:
length = len(array)
for max_int in range(0x1 << length):
subset = []
for i in range(length):
if max_int & (0x1 << i):
subset.append(array[i])
yield subset
and then it can be used as
def get_subsets(array):
powerset = []
for i in subsets(array):
powerser.append(i)
return powerset
Testing
Adding following in local file
if __name__ == '__main__':
sample = ['b', 'd', 'f']
for i in range(len(sample)):
print "Subsets for " , sample[i:], " are ", get_subsets(sample[i:])
gives following output
Subsets for ['b', 'd', 'f'] are [[], ['b'], ['d'], ['b', 'd'], ['f'], ['b', 'f'], ['d', 'f'], ['b', 'd', 'f']]
Subsets for ['d', 'f'] are [[], ['d'], ['f'], ['d', 'f']]
Subsets for ['f'] are [[], ['f']]
Almost all of these answers use list rather than set, which felt like a bit of a cheat to me. So, out of curiosity I tried to do a simple version truly on set and summarize for other "new to Python" folks.
I found there's a couple oddities in dealing with Python's set implementation. The main surprise to me was handling empty sets. This is in contrast to Ruby's Set implementation, where I can simply do Set[Set[]] and get a Set containing one empty Set, so I found it initially a little confusing.
To review, in doing powerset with sets, I encountered two problems:
set() takes an iterable, so set(set()) will return set() because the empty set iterable is empty (duh I guess :))
to get a set of sets, set({set()}) and set.add(set) won't work because set() isn't hashable
To solve both issues, I made use of frozenset(), which means I don't quite get what I want (type is literally set), but makes use of the overall set interace.
def powerset(original_set):
# below gives us a set with one empty set in it
ps = set({frozenset()})
for member in original_set:
subset = set()
for m in ps:
# to be added into subset, needs to be
# frozenset.union(set) so it's hashable
subset.add(m.union(set([member]))
ps = ps.union(subset)
return ps
Below we get 2² (16) frozensets correctly as output:
In [1]: powerset(set([1,2,3,4]))
Out[2]:
{frozenset(),
frozenset({3, 4}),
frozenset({2}),
frozenset({1, 4}),
frozenset({3}),
frozenset({2, 3}),
frozenset({2, 3, 4}),
frozenset({1, 2}),
frozenset({2, 4}),
frozenset({1}),
frozenset({1, 2, 4}),
frozenset({1, 3}),
frozenset({1, 2, 3}),
frozenset({4}),
frozenset({1, 3, 4}),
frozenset({1, 2, 3, 4})}
As there's no way to have a set of sets in Python, if you want to turn these frozensets into sets, you'll have to map them back into a list (list(map(set, powerset(set([1,2,3,4]))))
) or modify the above.
Perhaps the question is getting old, but I hope my code will help someone.
def powSet(set):
if len(set) == 0:
return [[]]
return addtoAll(set[0],powSet(set[1:])) + powSet(set[1:])
def addtoAll(e, set):
for c in set:
c.append(e)
return set
Getting all the subsets with recursion. Crazy-ass one-liner
from typing import List
def subsets(xs: list) -> List[list]:
return subsets(xs[1:]) + [x + [xs[0]] for x in subsets(xs[1:])] if xs else [[]]
Based on a Haskell solution
subsets :: [a] -> [[a]]
subsets [] = [[]]
subsets (x:xs) = map (x:) (subsets xs) ++ subsets xs
def findsubsets(s, n):
return list(itertools.combinations(s, n))
def allsubsets(s) :
a = []
for x in range(1,len(s)+1):
a.append(map(set,findsubsets(s,x)))
return a
def powerSet(s):
sets = [[]]
for i in s:
newsets = []
for k in sets:
newsets.append(k+[i])
sets += newsets
return sets
Code first, for those who want a simple answer.
I have a good explanation here https://leetcode.com/problems/subsets/solutions/3138042/simple-python-solution/
But the short answer is that you start with the set of the empty set, i.e. "sets = [[]]". I recommend to put a print a statement under "for i in s" i.e. "print(sets)" and see that it doubles for each element i
This is wild because none of these answers actually provide the return of an actual Python set. Here is a messy implementation that will give a powerset that actually is a Python set.
test_set = set(['yo', 'whatup', 'money'])
def powerset( base_set ):
""" modified from pydoc's itertools recipe shown above"""
from itertools import chain, combinations
base_list = list( base_set )
combo_list = [ combinations(base_list, r) for r in range(len(base_set)+1) ]
powerset = set([])
for ll in combo_list:
list_of_frozensets = list( map( frozenset, map( list, ll ) ) )
set_of_frozensets = set( list_of_frozensets )
powerset = powerset.union( set_of_frozensets )
return powerset
print powerset( test_set )
# >>> set([ frozenset(['money','whatup']), frozenset(['money','whatup','yo']),
# frozenset(['whatup']), frozenset(['whatup','yo']), frozenset(['yo']),
# frozenset(['money','yo']), frozenset(['money']), frozenset([]) ])
I'd love to see a better implementation, though.
Here is my quick implementation utilizing combinations but using only built-ins.
def powerSet(array):
length = str(len(array))
formatter = '{:0' + length + 'b}'
combinations = []
for i in xrange(2**int(length)):
combinations.append(formatter.format(i))
sets = set()
currentSet = []
for combo in combinations:
for i,val in enumerate(combo):
if val=='1':
currentSet.append(array[i])
sets.add(tuple(sorted(currentSet)))
currentSet = []
return sets
All subsets in range n as set:
n = int(input())
l = [i for i in range (1, n + 1)]
for number in range(2 ** n) :
binary = bin(number)[: 1 : -1]
subset = [l[i] for i in range(len(binary)) if binary[i] == "1"]
print(set(sorted(subset)) if number > 0 else "{}")
import math
def printPowerSet(set,set_size):
pow_set_size =int(math.pow(2, set_size))
for counter in range(pow_set_size):
for j in range(set_size):
if((counter & (1 << j)) > 0):
print(set[j], end = "")
print("")
set = ['a', 'b', 'c']
printPowerSet(set,3)
A variation of the question, is an exercise I see on the book "Discovering Computer Science: Interdisciplinary Problems, Principles, and Python Programming. 2015 edition". In that exercise 10.2.11, the input is just an integer number, and the output should be the power sets. Here is my recursive solution (not using anything else but basic python3 )
def powerSetR(n):
assert n >= 0
if n == 0:
return [[]]
else:
input_set = list(range(1, n+1)) # [1,2,...n]
main_subset = [ ]
for small_subset in powerSetR(n-1):
main_subset += [small_subset]
main_subset += [ [input_set[-1]] + small_subset]
return main_subset
superset = powerSetR(4)
print(superset)
print("Number of sublists:", len(superset))
And the output is
[[], [4], [3], [4, 3], [2], [4, 2], [3, 2], [4, 3, 2], [1], [4, 1], [3, 1], [4, 3, 1], [2, 1], [4, 2, 1], [3, 2, 1], [4, 3, 2, 1]]
Number of sublists: 16
I hadn't come across the more_itertools.powerset function and would recommend using that. I also recommend not using the default ordering of the output from itertools.combinations, often instead you want to minimise the distance between the positions and sort the subsets of items with shorter distance between them above/before the items with larger distance between them.
The itertools recipes page shows it uses chain.from_iterable
Note that the r here matches the standard notation for the lower part of a binomial coefficient, the s is usually referred to as n in mathematics texts and on calculators (“n Choose r”)
def powerset(iterable):
"powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)"
s = list(iterable)
return chain.from_iterable(combinations(s, r) for r in range(len(s)+1))
The other examples here give the powerset of [1,2,3,4] in such a way that the 2-tuples are listed in "lexicographic" order (when we print the numbers as integers). If I write the distance between the numbers alongside it (i.e. the difference), it shows my point:
12 ⇒ 1
13 ⇒ 2
14 ⇒ 3
23 ⇒ 1
24 ⇒ 2
34 ⇒ 1
The correct order for subsets should be the order which 'exhausts' the minimal distance first, like so:
12 ⇒ 1
23 ⇒ 1
34 ⇒ 1
13 ⇒ 2
24 ⇒ 2
14 ⇒ 3
Using numbers here makes this ordering look 'wrong', but consider for example the letters ["a","b","c","d"] it is clearer why this might be useful to obtain the powerset in this order:
ab ⇒ 1
bc ⇒ 1
cd ⇒ 1
ac ⇒ 2
bd ⇒ 2
ad ⇒ 3
This effect is more pronounced with more items, and for my purposes it makes the difference between being able to describe the ranges of the indexes of the powerset meaningfully.
(There is a lot written on Gray codes etc. for the output order of algorithms in combinatorics, I don't see it as a side issue).
I actually just wrote a fairly involved program which used this fast integer partition code to output the values in the proper order, but then I discovered more_itertools.powerset and for most uses it's probably fine to just use that function like so:
from more_itertools import powerset
from numpy import ediff1d
def ps_sorter(tup):
l = len(tup)
d = ediff1d(tup).tolist()
return l, d
ps = powerset([1,2,3,4])
ps = sorted(ps, key=ps_sorter)
for x in ps:
print(x)
⇣
()
(1,)
(2,)
(3,)
(4,)
(1, 2)
(2, 3)
(3, 4)
(1, 3)
(2, 4)
(1, 4)
(1, 2, 3)
(2, 3, 4)
(1, 2, 4)
(1, 3, 4)
(1, 2, 3, 4)
I wrote some more involved code which will print the powerset nicely (see the repo for pretty printing functions I've not included here: print_partitions, print_partitions_by_length, and pprint_tuple).
Repo: ordered-powerset, specifically pset_partitions.py
This is all pretty simple, but still might be useful if you want some code that'll let you get straight to accessing the different levels of the powerset:
from itertools import permutations as permute
from numpy import cumsum
# http://jeromekelleher.net/generating-integer-partitions.html
# via
# https://stackoverflow.com/questions/10035752/elegant-python-code-for-integer-partitioning#comment25080713_10036764
def asc_int_partitions(n):
a = [0 for i in range(n + 1)]
k = 1
y = n - 1
while k != 0:
x = a[k - 1] + 1
k -= 1
while 2 * x <= y:
a[k] = x
y -= x
k += 1
l = k + 1
while x <= y:
a[k] = x
a[l] = y
yield tuple(a[:k + 2])
x += 1
y -= 1
a[k] = x + y
y = x + y - 1
yield tuple(a[:k + 1])
# https://stackoverflow.com/a/6285330/2668831
def uniquely_permute(iterable, enforce_sort=False, r=None):
previous = tuple()
if enforce_sort: # potential waste of effort (default: False)
iterable = sorted(iterable)
for p in permute(iterable, r):
if p > previous:
previous = p
yield p
def sum_min(p):
return sum(p), min(p)
def partitions_by_length(max_n, sorting=True, permuting=False):
partition_dict = {0: ()}
for n in range(1,max_n+1):
partition_dict.setdefault(n, [])
partitions = list(asc_int_partitions(n))
for p in partitions:
if permuting:
perms = uniquely_permute(p)
for perm in perms:
partition_dict.get(len(p)).append(perm)
else:
partition_dict.get(len(p)).append(p)
if not sorting:
return partition_dict
for k in partition_dict:
partition_dict.update({k: sorted(partition_dict.get(k), key=sum_min)})
return partition_dict
def print_partitions_by_length(max_n, sorting=True, permuting=True):
partition_dict = partitions_by_length(max_n, sorting=sorting, permuting=permuting)
for k in partition_dict:
if k == 0:
print(tuple(partition_dict.get(k)), end="")
for p in partition_dict.get(k):
print(pprint_tuple(p), end=" ")
print()
return
def generate_powerset(items, subset_handler=tuple, verbose=False):
"""
Generate the powerset of an iterable `items`.
Handling of the elements of the iterable is by whichever function is passed as
`subset_handler`, which must be able to handle the `None` value for the
empty set. The function `string_handler` will join the elements of the subset
with the empty string (useful when `items` is an iterable of `str` variables).
"""
ps = {0: [subset_handler()]}
n = len(items)
p_dict = partitions_by_length(n-1, sorting=True, permuting=True)
for p_len, parts in p_dict.items():
ps.setdefault(p_len, [])
if p_len == 0:
# singletons
for offset in range(n):
subset = subset_handler([items[offset]])
if verbose:
if offset > 0:
print(end=" ")
if offset == n - 1:
print(subset, end="\n")
else:
print(subset, end=",")
ps.get(p_len).append(subset)
for pcount, partition in enumerate(parts):
distance = sum(partition)
indices = (cumsum(partition)).tolist()
for offset in range(n - distance):
subset = subset_handler([items[offset]] + [items[offset:][i] for i in indices])
if verbose:
if offset > 0:
print(end=" ")
if offset == n - distance - 1:
print(subset, end="\n")
else:
print(subset, end=",")
ps.get(p_len).append(subset)
if verbose and p_len < n-1:
print()
return ps
As an example, I wrote a CLI demo program which takes a string as a command line argument:
python string_powerset.py abcdef
⇣
a, b, c, d, e, f
ab, bc, cd, de, ef
ac, bd, ce, df
ad, be, cf
ae, bf
af
abc, bcd, cde, def
abd, bce, cdf
acd, bde, cef
abe, bcf
ade, bef
ace, bdf
abf
aef
acf
adf
abcd, bcde, cdef
abce, bcdf
abde, bcef
acde, bdef
abcf
abef
adef
abdf
acdf
acef
abcde, bcdef
abcdf
abcef
abdef
acdef
abcdef
Here it is my solutions, it is similar (conceptually) with the solution of lmiguelvargasf.
Let me say that
-[math item] by defintion the powerset do contain the empty set
-[personal taste] and also that I don't like using frozenset.
So the input is a list and the output will be a list of lists. The function could close earlier, but I like the element of the power set to be order lexicographically, that essentially means nicely.
def power_set(L):
"""
L is a list.
The function returns the power set, but as a list of lists.
"""
cardinality=len(L)
n=2 ** cardinality
powerset = []
for i in range(n):
a=bin(i)[2:]
subset=[]
for j in range(len(a)):
if a[-j-1]=='1':
subset.append(L[j])
powerset.append(subset)
#the function could stop here closing with
#return powerset
powerset_orderred=[]
for k in range(cardinality+1):
for w in powerset:
if len(w)==k:
powerset_orderred.append(w)
return powerset_orderred
I am using Python to learn linear algebra and I have two dictionaries:
v = {1: 1, 2: 8, 3: 0}
and
M = {(1, 2): 2, (3, 1): 1, (3, 3): 7, (2, 1): -1}
and I want to make a dictionary that adds all elements where the keys in v is the same as the first part of the tuples in M. Example of what I want as a answer for the two dictionaries here. (I will show the calculations I want to do)
newDict = {1: 1*M[(1, 2)], 2: 8*M[(2, 1)], 3: 0*M[(3, 1)]+0*M[(3, 3)]
which is the same as:
newDict = {1: 1*2, 2: 8*-1, 3: 0*1+3*7}
so I get a final dictionary in the form
newDict = {1:2, 2:-8, 3:0}
as you can see, I want the same keys as in the dictionary v. The closest I have gotten is this:
>>> [v[k]*M[r] for k in v for r in M if k == r[0]]
[2, -8]
Where I at least have the right answers, but I cannot get this to work. I don't know where to go from here or if I am at the right track at all. Sorry if my explanation might be lacking
Because you are basing values on multiple input keys, use a loop, not a comprehension. Using a collections.defaultdict object makes the logic a little simpler too:
from collections import defaultdict
newDict = defaultdict(int)
for x, y in M:
newDict[x] += M[x, y] * v.get(x, 0)
Output:
>>> from collections import defaultdict
>>> v = {1: 1, 2: 8, 3: 0}
>>> M = {(1, 2): 2, (3, 1): 1, (3, 3): 7, (2, 1): -1}
>>> newDict = defaultdict(int)
>>> for x, y in M:
... newDict[x] += M[x, y] * v.get(x, 0)
...
>>> newDict
defaultdict(<type 'int'>, {1: 2, 2: -8, 3: 0})
How about this..
newD = {k : 0 for k in v}
for k in v:
for r in M:
if k == r[0]: newD[k] += v[k]*M[r]
Given a set
{0, 1, 2, 3}
How can I produce the subsets:
[set(),
{0},
{1},
{2},
{3},
{0, 1},
{0, 2},
{0, 3},
{1, 2},
{1, 3},
{2, 3},
{0, 1, 2},
{0, 1, 3},
{0, 2, 3},
{1, 2, 3},
{0, 1, 2, 3}]
The Python itertools page has exactly a powerset recipe for this:
from itertools import chain, combinations
def powerset(iterable):
"powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)"
s = list(iterable)
return chain.from_iterable(combinations(s, r) for r in range(len(s)+1))
Output:
>>> list(powerset("abcd"))
[(), ('a',), ('b',), ('c',), ('d',), ('a', 'b'), ('a', 'c'), ('a', 'd'), ('b', 'c'), ('b', 'd'), ('c', 'd'), ('a', 'b', 'c'), ('a', 'b', 'd'), ('a', 'c', 'd'), ('b', 'c', 'd'), ('a', 'b', 'c', 'd')]
If you don't like that empty tuple at the beginning, you can just change the range statement to range(1, len(s)+1) to avoid a 0-length combination.
Here is more code for a powerset. This is written from scratch:
>>> def powerset(s):
... x = len(s)
... for i in range(1 << x):
... print [s[j] for j in range(x) if (i & (1 << j))]
...
>>> powerset([4,5,6])
[]
[4]
[5]
[4, 5]
[6]
[4, 6]
[5, 6]
[4, 5, 6]
Mark Rushakoff's comment is applicable here: "If you don't like that empty tuple at the beginning, on."you can just change the range statement to range(1, len(s)+1) to avoid a 0-length combination", except in my case you change for i in range(1 << x) to for i in range(1, 1 << x).
Returning to this years later, I'd now write it like this:
def powerset(s):
x = len(s)
masks = [1 << i for i in range(x)]
for i in range(1 << x):
yield [ss for mask, ss in zip(masks, s) if i & mask]
And then the test code would look like this, say:
print(list(powerset([4, 5, 6])))
Using yield means that you do not need to calculate all results in a single piece of memory. Precalculating the masks outside the main loop is assumed to be a worthwhile optimization.
If you're looking for a quick answer, I just searched "python power set" on google and came up with this: Python Power Set Generator
Here's a copy-paste from the code in that page:
def powerset(seq):
"""
Returns all the subsets of this set. This is a generator.
"""
if len(seq) <= 1:
yield seq
yield []
else:
for item in powerset(seq[1:]):
yield [seq[0]]+item
yield item
This can be used like this:
l = [1, 2, 3, 4]
r = [x for x in powerset(l)]
Now r is a list of all the elements you wanted, and can be sorted and printed:
r.sort()
print r
[[], [1], [1, 2], [1, 2, 3], [1, 2, 3, 4], [1, 2, 4], [1, 3], [1, 3, 4], [1, 4], [2], [2, 3], [2, 3, 4], [2, 4], [3], [3, 4], [4]]
Use function powerset() from package more_itertools.
Yields all possible subsets of the iterable
>>> list(powerset([1, 2, 3]))
[(), (1,), (2,), (3,), (1, 2), (1, 3), (2, 3), (1, 2, 3)]
If you want sets, use:
list(map(set, powerset(iterable)))
I have found the following algorithm very clear and simple:
def get_powerset(some_list):
"""Returns all subsets of size 0 - len(some_list) for some_list"""
if len(some_list) == 0:
return [[]]
subsets = []
first_element = some_list[0]
remaining_list = some_list[1:]
# Strategy: get all the subsets of remaining_list. For each
# of those subsets, a full subset list will contain both
# the original subset as well as a version of the subset
# that contains first_element
for partial_subset in get_powerset(remaining_list):
subsets.append(partial_subset)
subsets.append(partial_subset[:] + [first_element])
return subsets
Another way one can generate the powerset is by generating all binary numbers that have n bits. As a power set the amount of number with n digits is 2 ^ n. The principle of this algorithm is that an element could be present or not in a subset as a binary digit could be one or zero but not both.
def power_set(items):
N = len(items)
# enumerate the 2 ** N possible combinations
for i in range(2 ** N):
combo = []
for j in range(N):
# test bit jth of integer i
if (i >> j) % 2 == 1:
combo.append(items[j])
yield combo
I found both algorithms when I was taking MITx: 6.00.2x Introduction to Computational Thinking and Data Science, and I consider it is one of the easiest algorithms to understand I have seen.
from functools import reduce
def powerset(lst):
return reduce(lambda result, x: result + [subset + [x] for subset in result],
lst, [[]])
There is a refinement of powerset:
def powerset(seq):
"""
Returns all the subsets of this set. This is a generator.
"""
if len(seq) <= 0:
yield []
else:
for item in powerset(seq[1:]):
yield [seq[0]]+item
yield item
TL;DR (go directly to Simplification)
I know I have previously added an answer, but I really like my new implementation. I am taking a set as input, but it actually could be any iterable, and I am returning a set of sets which is the power set of the input. I like this approach because it is more aligned with the mathematical definition of power set (set of all subsets).
def power_set(A):
"""A is an iterable (list, tuple, set, str, etc)
returns a set which is the power set of A."""
length = len(A)
l = [a for a in A]
ps = set()
for i in range(2 ** length):
selector = f'{i:0{length}b}'
subset = {l[j] for j, bit in enumerate(selector) if bit == '1'}
ps.add(frozenset(subset))
return ps
If you want exactly the output you posted in your answer use this:
>>> [set(s) for s in power_set({1, 2, 3, 4})]
[{3, 4},
{2},
{1, 4},
{2, 3, 4},
{2, 3},
{1, 2, 4},
{1, 2},
{1, 2, 3},
{3},
{2, 4},
{1},
{1, 2, 3, 4},
set(),
{1, 3},
{1, 3, 4},
{4}]
Explanation
It is known that the number of elements of the power set is 2 ** len(A), so that could clearly be seen in the for loop.
I need to convert the input (ideally a set) into a list because by a set is a data structure of unique unordered elements, and the order will be crucial to generate the subsets.
selector is key in this algorithm. Note that selector has the same length as the input set, and to make this possible it is using an f-string with padding. Basically, this allows me to select the elements that will be added to each subset during each iteration. Let's say the input set has 3 elements {0, 1, 2}, so selector will take values between 0 and 7 (inclusive), which in binary are:
000 # 0
001 # 1
010 # 2
011 # 3
100 # 4
101 # 5
110 # 6
111 # 7
So, each bit could serve as an indicator if an element of the original set should be added or not. Look at the binary numbers, and just think of each number as an element of the super set in which 1 means that an element at index j should be added, and 0 means that this element should not be added.
I am using a set comprehension to generate a subset at each iteration, and I convert this subset into a frozenset so I can add it to ps (power set). Otherwise, I won't be able to add it because a set in Python consists only of immutable objects.
Simplification
You can simplify the code using some python comprehensions, so you can get rid of those for loops. You can also use zip to avoid using j index and the code will end up as the following:
def power_set(A):
length = len(A)
return {
frozenset({e for e, b in zip(A, f'{i:{length}b}') if b == '1'})
for i in range(2 ** length)
}
That's it. What I like of this algorithm is that is clearer and more intuitive than others because it looks quite magical to rely on itertools even though it works as expected.
def get_power_set(s):
power_set=[[]]
for elem in s:
# iterate over the sub sets so far
for sub_set in power_set:
# add a new subset consisting of the subset at hand added elem to it
# effectively doubling the sets resulting in the 2^n sets in the powerset of s.
power_set=power_set+[list(sub_set)+[elem]]
return power_set
For example:
get_power_set([1,2,3])
yield
[[], [1], [2], [1, 2], [3], [1, 3], [2, 3], [1, 2, 3]]
This can be done very naturally with itertools.product:
import itertools
def powerset(l):
for sl in itertools.product(*[[[], [i]] for i in l]):
yield {j for i in sl for j in i}
I know this is too late
There are many other solutions already but still...
def power_set(lst):
pw_set = [[]]
for i in range(0,len(lst)):
for j in range(0,len(pw_set)):
ele = pw_set[j].copy()
ele = ele + [lst[i]]
pw_set = pw_set + [ele]
return pw_set
I just wanted to provide the most comprehensible solution, the anti code-golf version.
from itertools import combinations
l = ["x", "y", "z", ]
def powerset(items):
combo = []
for r in range(len(items) + 1):
#use a list to coerce a actual list from the combinations generator
combo.append(list(combinations(items,r)))
return combo
l_powerset = powerset(l)
for i, item in enumerate(l_powerset):
print "All sets of length ", i
print item
The results
All sets of length 0
[()]
All sets of length 1
[('x',), ('y',), ('z',)]
All sets of length 2
[('x', 'y'), ('x', 'z'), ('y', 'z')]
All sets of length 3
[('x', 'y', 'z')]
For more see the itertools docs, also the wikipedia entry on power sets
With empty set, which is part of all the subsets, you could use:
def subsets(iterable):
for n in range(len(iterable) + 1):
yield from combinations(iterable, n)
from itertools import combinations
def subsets(arr: set) -> list:
subsets = []
[subsets.extend(list(combinations(arr,n))) for n in range(len(arr))]
return subsets
a = {1,2,3}
print(subsets(a))
Output:
[(), (1,), (2,), (3,), (1, 2), (1, 3), (2, 3)]
For sorted subsets, we can do:
# sorted subsets
print(sorted(subsets(a)))
Output:
[(), (1,), (1, 2), (1, 3), (2,), (2, 3), (3,)]
Just a quick power set refresher !
Power set of a set X, is simply the set of all subsets of X including
the empty set
Example set X = (a,b,c)
Power Set = { { a , b , c } , { a , b } , { a , c } , { b , c } , { a } , { b } , { c } , { } }
Here is another way of finding power set:
def power_set(input):
# returns a list of all subsets of the list a
if (len(input) == 0):
return [[]]
else:
main_subset = [ ]
for small_subset in power_set(input[1:]):
main_subset += [small_subset]
main_subset += [[input[0]] + small_subset]
return main_subset
print(power_set([0,1,2,3]))
full credit to source
If you want any specific length of subsets you can do it like this:
from itertools import combinations
someSet = {0, 1, 2, 3}
([x for i in range(len(someSet)+1) for x in combinations(someSet,i)])
More generally for arbitary length subsets you can modify the range arugment. The output is
[(), (0,), (1,), (2,), (3,), (0, 1), (0, 2), (0, 3), (1, 2), (1, 3), (2, 3), (0, 1, 2), (0, 1, 3), (0, 2, 3), (1, 2, 3), (0, 1, 2, 3)]
You can do it like this:
def powerset(x):
m=[]
if not x:
m.append(x)
else:
A = x[0]
B = x[1:]
for z in powerset(B):
m.append(z)
r = [A] + z
m.append(r)
return m
print(powerset([1, 2, 3, 4]))
Output:
[[], [1], [2], [1, 2], [3], [1, 3], [2, 3], [1, 2, 3], [4], [1, 4], [2, 4], [1, 2, 4], [3, 4], [1, 3, 4], [2, 3, 4], [1, 2, 3, 4]]
A simple way would be to harness the internal representation of integers under 2's complement arithmetic.
Binary representation of integers is as {000, 001, 010, 011, 100, 101, 110, 111} for numbers ranging from 0 to 7. For an integer counter value, considering 1 as inclusion of corresponding element in collection and '0' as exclusion we can generate subsets based on the counting sequence. Numbers have to be generated from 0 to pow(2,n) -1 where n is the length of array i.e. number of bits in binary representation.
A simple Subset Generator Function based on it can be written as below. It basically relies
def subsets(array):
if not array:
return
else:
length = len(array)
for max_int in range(0x1 << length):
subset = []
for i in range(length):
if max_int & (0x1 << i):
subset.append(array[i])
yield subset
and then it can be used as
def get_subsets(array):
powerset = []
for i in subsets(array):
powerser.append(i)
return powerset
Testing
Adding following in local file
if __name__ == '__main__':
sample = ['b', 'd', 'f']
for i in range(len(sample)):
print "Subsets for " , sample[i:], " are ", get_subsets(sample[i:])
gives following output
Subsets for ['b', 'd', 'f'] are [[], ['b'], ['d'], ['b', 'd'], ['f'], ['b', 'f'], ['d', 'f'], ['b', 'd', 'f']]
Subsets for ['d', 'f'] are [[], ['d'], ['f'], ['d', 'f']]
Subsets for ['f'] are [[], ['f']]
Almost all of these answers use list rather than set, which felt like a bit of a cheat to me. So, out of curiosity I tried to do a simple version truly on set and summarize for other "new to Python" folks.
I found there's a couple oddities in dealing with Python's set implementation. The main surprise to me was handling empty sets. This is in contrast to Ruby's Set implementation, where I can simply do Set[Set[]] and get a Set containing one empty Set, so I found it initially a little confusing.
To review, in doing powerset with sets, I encountered two problems:
set() takes an iterable, so set(set()) will return set() because the empty set iterable is empty (duh I guess :))
to get a set of sets, set({set()}) and set.add(set) won't work because set() isn't hashable
To solve both issues, I made use of frozenset(), which means I don't quite get what I want (type is literally set), but makes use of the overall set interace.
def powerset(original_set):
# below gives us a set with one empty set in it
ps = set({frozenset()})
for member in original_set:
subset = set()
for m in ps:
# to be added into subset, needs to be
# frozenset.union(set) so it's hashable
subset.add(m.union(set([member]))
ps = ps.union(subset)
return ps
Below we get 2² (16) frozensets correctly as output:
In [1]: powerset(set([1,2,3,4]))
Out[2]:
{frozenset(),
frozenset({3, 4}),
frozenset({2}),
frozenset({1, 4}),
frozenset({3}),
frozenset({2, 3}),
frozenset({2, 3, 4}),
frozenset({1, 2}),
frozenset({2, 4}),
frozenset({1}),
frozenset({1, 2, 4}),
frozenset({1, 3}),
frozenset({1, 2, 3}),
frozenset({4}),
frozenset({1, 3, 4}),
frozenset({1, 2, 3, 4})}
As there's no way to have a set of sets in Python, if you want to turn these frozensets into sets, you'll have to map them back into a list (list(map(set, powerset(set([1,2,3,4]))))
) or modify the above.
Perhaps the question is getting old, but I hope my code will help someone.
def powSet(set):
if len(set) == 0:
return [[]]
return addtoAll(set[0],powSet(set[1:])) + powSet(set[1:])
def addtoAll(e, set):
for c in set:
c.append(e)
return set
Getting all the subsets with recursion. Crazy-ass one-liner
from typing import List
def subsets(xs: list) -> List[list]:
return subsets(xs[1:]) + [x + [xs[0]] for x in subsets(xs[1:])] if xs else [[]]
Based on a Haskell solution
subsets :: [a] -> [[a]]
subsets [] = [[]]
subsets (x:xs) = map (x:) (subsets xs) ++ subsets xs
def findsubsets(s, n):
return list(itertools.combinations(s, n))
def allsubsets(s) :
a = []
for x in range(1,len(s)+1):
a.append(map(set,findsubsets(s,x)))
return a
def powerSet(s):
sets = [[]]
for i in s:
newsets = []
for k in sets:
newsets.append(k+[i])
sets += newsets
return sets
Code first, for those who want a simple answer.
I have a good explanation here https://leetcode.com/problems/subsets/solutions/3138042/simple-python-solution/
But the short answer is that you start with the set of the empty set, i.e. "sets = [[]]". I recommend to put a print a statement under "for i in s" i.e. "print(sets)" and see that it doubles for each element i
This is wild because none of these answers actually provide the return of an actual Python set. Here is a messy implementation that will give a powerset that actually is a Python set.
test_set = set(['yo', 'whatup', 'money'])
def powerset( base_set ):
""" modified from pydoc's itertools recipe shown above"""
from itertools import chain, combinations
base_list = list( base_set )
combo_list = [ combinations(base_list, r) for r in range(len(base_set)+1) ]
powerset = set([])
for ll in combo_list:
list_of_frozensets = list( map( frozenset, map( list, ll ) ) )
set_of_frozensets = set( list_of_frozensets )
powerset = powerset.union( set_of_frozensets )
return powerset
print powerset( test_set )
# >>> set([ frozenset(['money','whatup']), frozenset(['money','whatup','yo']),
# frozenset(['whatup']), frozenset(['whatup','yo']), frozenset(['yo']),
# frozenset(['money','yo']), frozenset(['money']), frozenset([]) ])
I'd love to see a better implementation, though.
Here is my quick implementation utilizing combinations but using only built-ins.
def powerSet(array):
length = str(len(array))
formatter = '{:0' + length + 'b}'
combinations = []
for i in xrange(2**int(length)):
combinations.append(formatter.format(i))
sets = set()
currentSet = []
for combo in combinations:
for i,val in enumerate(combo):
if val=='1':
currentSet.append(array[i])
sets.add(tuple(sorted(currentSet)))
currentSet = []
return sets
All subsets in range n as set:
n = int(input())
l = [i for i in range (1, n + 1)]
for number in range(2 ** n) :
binary = bin(number)[: 1 : -1]
subset = [l[i] for i in range(len(binary)) if binary[i] == "1"]
print(set(sorted(subset)) if number > 0 else "{}")
import math
def printPowerSet(set,set_size):
pow_set_size =int(math.pow(2, set_size))
for counter in range(pow_set_size):
for j in range(set_size):
if((counter & (1 << j)) > 0):
print(set[j], end = "")
print("")
set = ['a', 'b', 'c']
printPowerSet(set,3)
A variation of the question, is an exercise I see on the book "Discovering Computer Science: Interdisciplinary Problems, Principles, and Python Programming. 2015 edition". In that exercise 10.2.11, the input is just an integer number, and the output should be the power sets. Here is my recursive solution (not using anything else but basic python3 )
def powerSetR(n):
assert n >= 0
if n == 0:
return [[]]
else:
input_set = list(range(1, n+1)) # [1,2,...n]
main_subset = [ ]
for small_subset in powerSetR(n-1):
main_subset += [small_subset]
main_subset += [ [input_set[-1]] + small_subset]
return main_subset
superset = powerSetR(4)
print(superset)
print("Number of sublists:", len(superset))
And the output is
[[], [4], [3], [4, 3], [2], [4, 2], [3, 2], [4, 3, 2], [1], [4, 1], [3, 1], [4, 3, 1], [2, 1], [4, 2, 1], [3, 2, 1], [4, 3, 2, 1]]
Number of sublists: 16
I hadn't come across the more_itertools.powerset function and would recommend using that. I also recommend not using the default ordering of the output from itertools.combinations, often instead you want to minimise the distance between the positions and sort the subsets of items with shorter distance between them above/before the items with larger distance between them.
The itertools recipes page shows it uses chain.from_iterable
Note that the r here matches the standard notation for the lower part of a binomial coefficient, the s is usually referred to as n in mathematics texts and on calculators (“n Choose r”)
def powerset(iterable):
"powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)"
s = list(iterable)
return chain.from_iterable(combinations(s, r) for r in range(len(s)+1))
The other examples here give the powerset of [1,2,3,4] in such a way that the 2-tuples are listed in "lexicographic" order (when we print the numbers as integers). If I write the distance between the numbers alongside it (i.e. the difference), it shows my point:
12 ⇒ 1
13 ⇒ 2
14 ⇒ 3
23 ⇒ 1
24 ⇒ 2
34 ⇒ 1
The correct order for subsets should be the order which 'exhausts' the minimal distance first, like so:
12 ⇒ 1
23 ⇒ 1
34 ⇒ 1
13 ⇒ 2
24 ⇒ 2
14 ⇒ 3
Using numbers here makes this ordering look 'wrong', but consider for example the letters ["a","b","c","d"] it is clearer why this might be useful to obtain the powerset in this order:
ab ⇒ 1
bc ⇒ 1
cd ⇒ 1
ac ⇒ 2
bd ⇒ 2
ad ⇒ 3
This effect is more pronounced with more items, and for my purposes it makes the difference between being able to describe the ranges of the indexes of the powerset meaningfully.
(There is a lot written on Gray codes etc. for the output order of algorithms in combinatorics, I don't see it as a side issue).
I actually just wrote a fairly involved program which used this fast integer partition code to output the values in the proper order, but then I discovered more_itertools.powerset and for most uses it's probably fine to just use that function like so:
from more_itertools import powerset
from numpy import ediff1d
def ps_sorter(tup):
l = len(tup)
d = ediff1d(tup).tolist()
return l, d
ps = powerset([1,2,3,4])
ps = sorted(ps, key=ps_sorter)
for x in ps:
print(x)
⇣
()
(1,)
(2,)
(3,)
(4,)
(1, 2)
(2, 3)
(3, 4)
(1, 3)
(2, 4)
(1, 4)
(1, 2, 3)
(2, 3, 4)
(1, 2, 4)
(1, 3, 4)
(1, 2, 3, 4)
I wrote some more involved code which will print the powerset nicely (see the repo for pretty printing functions I've not included here: print_partitions, print_partitions_by_length, and pprint_tuple).
Repo: ordered-powerset, specifically pset_partitions.py
This is all pretty simple, but still might be useful if you want some code that'll let you get straight to accessing the different levels of the powerset:
from itertools import permutations as permute
from numpy import cumsum
# http://jeromekelleher.net/generating-integer-partitions.html
# via
# https://stackoverflow.com/questions/10035752/elegant-python-code-for-integer-partitioning#comment25080713_10036764
def asc_int_partitions(n):
a = [0 for i in range(n + 1)]
k = 1
y = n - 1
while k != 0:
x = a[k - 1] + 1
k -= 1
while 2 * x <= y:
a[k] = x
y -= x
k += 1
l = k + 1
while x <= y:
a[k] = x
a[l] = y
yield tuple(a[:k + 2])
x += 1
y -= 1
a[k] = x + y
y = x + y - 1
yield tuple(a[:k + 1])
# https://stackoverflow.com/a/6285330/2668831
def uniquely_permute(iterable, enforce_sort=False, r=None):
previous = tuple()
if enforce_sort: # potential waste of effort (default: False)
iterable = sorted(iterable)
for p in permute(iterable, r):
if p > previous:
previous = p
yield p
def sum_min(p):
return sum(p), min(p)
def partitions_by_length(max_n, sorting=True, permuting=False):
partition_dict = {0: ()}
for n in range(1,max_n+1):
partition_dict.setdefault(n, [])
partitions = list(asc_int_partitions(n))
for p in partitions:
if permuting:
perms = uniquely_permute(p)
for perm in perms:
partition_dict.get(len(p)).append(perm)
else:
partition_dict.get(len(p)).append(p)
if not sorting:
return partition_dict
for k in partition_dict:
partition_dict.update({k: sorted(partition_dict.get(k), key=sum_min)})
return partition_dict
def print_partitions_by_length(max_n, sorting=True, permuting=True):
partition_dict = partitions_by_length(max_n, sorting=sorting, permuting=permuting)
for k in partition_dict:
if k == 0:
print(tuple(partition_dict.get(k)), end="")
for p in partition_dict.get(k):
print(pprint_tuple(p), end=" ")
print()
return
def generate_powerset(items, subset_handler=tuple, verbose=False):
"""
Generate the powerset of an iterable `items`.
Handling of the elements of the iterable is by whichever function is passed as
`subset_handler`, which must be able to handle the `None` value for the
empty set. The function `string_handler` will join the elements of the subset
with the empty string (useful when `items` is an iterable of `str` variables).
"""
ps = {0: [subset_handler()]}
n = len(items)
p_dict = partitions_by_length(n-1, sorting=True, permuting=True)
for p_len, parts in p_dict.items():
ps.setdefault(p_len, [])
if p_len == 0:
# singletons
for offset in range(n):
subset = subset_handler([items[offset]])
if verbose:
if offset > 0:
print(end=" ")
if offset == n - 1:
print(subset, end="\n")
else:
print(subset, end=",")
ps.get(p_len).append(subset)
for pcount, partition in enumerate(parts):
distance = sum(partition)
indices = (cumsum(partition)).tolist()
for offset in range(n - distance):
subset = subset_handler([items[offset]] + [items[offset:][i] for i in indices])
if verbose:
if offset > 0:
print(end=" ")
if offset == n - distance - 1:
print(subset, end="\n")
else:
print(subset, end=",")
ps.get(p_len).append(subset)
if verbose and p_len < n-1:
print()
return ps
As an example, I wrote a CLI demo program which takes a string as a command line argument:
python string_powerset.py abcdef
⇣
a, b, c, d, e, f
ab, bc, cd, de, ef
ac, bd, ce, df
ad, be, cf
ae, bf
af
abc, bcd, cde, def
abd, bce, cdf
acd, bde, cef
abe, bcf
ade, bef
ace, bdf
abf
aef
acf
adf
abcd, bcde, cdef
abce, bcdf
abde, bcef
acde, bdef
abcf
abef
adef
abdf
acdf
acef
abcde, bcdef
abcdf
abcef
abdef
acdef
abcdef
Here it is my solutions, it is similar (conceptually) with the solution of lmiguelvargasf.
Let me say that
-[math item] by defintion the powerset do contain the empty set
-[personal taste] and also that I don't like using frozenset.
So the input is a list and the output will be a list of lists. The function could close earlier, but I like the element of the power set to be order lexicographically, that essentially means nicely.
def power_set(L):
"""
L is a list.
The function returns the power set, but as a list of lists.
"""
cardinality=len(L)
n=2 ** cardinality
powerset = []
for i in range(n):
a=bin(i)[2:]
subset=[]
for j in range(len(a)):
if a[-j-1]=='1':
subset.append(L[j])
powerset.append(subset)
#the function could stop here closing with
#return powerset
powerset_orderred=[]
for k in range(cardinality+1):
for w in powerset:
if len(w)==k:
powerset_orderred.append(w)
return powerset_orderred