我正在尝试为Python找到或开发Integer Partitioning代码。
FYI,整数分区将给定的整数n表示为小于n的整数之和。例如,整数5可以表示为4 + 1 = 3 + 2 = 3 + 1 + 1 = 2 + 2 + 1 = 2 + 1 + 1 + 1 = 1 + 1 + 1 + 1 + 1
我已经为此找到了许多解决方案。 http://homepages.ed.ac.uk/jkellehe/partitions.php和http://code.activestate.com/recipes/218332-generator-for-integer-partitions/
但是,我真正想要的是限制分区的数量。
说,分区#em> k = 2,程序只需显示5 = 4 + 1 = 3 + 2
,
如果 k = 3,5 = 3 + 1 + 1 = 2 + 2 + 1
答案 0 :(得分:19)
我写了一个生成器解决方案
def partitionfunc(n,k,l=1):
'''n is the integer to partition, k is the length of partitions, l is the min partition element size'''
if k < 1:
raise StopIteration
if k == 1:
if n >= l:
yield (n,)
raise StopIteration
for i in range(l,n+1):
for result in partitionfunc(n-i,k-1,i):
yield (i,)+result
这将生成长度为n
的{{1}}的所有分区,每个分区的顺序最小到最大。
快速说明:通过k
,使用生成器方法似乎比使用falsetru的直接方法快得多,使用测试函数cProfile
。在lambda x,y: list(partitionfunc(x,y))
的测试运行中,我的代码在.019秒内与直接方法的2.612秒相比。
答案 1 :(得分:4)
def part(n, k):
def _part(n, k, pre):
if n <= 0:
return []
if k == 1:
if n <= pre:
return [[n]]
return []
ret = []
for i in range(min(pre, n), 0, -1):
ret += [[i] + sub for sub in _part(n-i, k-1, i)]
return ret
return _part(n, k, n)
示例:
>>> part(5, 1)
[[5]]
>>> part(5, 2)
[[4, 1], [3, 2]]
>>> part(5, 3)
[[3, 1, 1], [2, 2, 1]]
>>> part(5, 4)
[[2, 1, 1, 1]]
>>> part(5, 5)
[[1, 1, 1, 1, 1]]
>>> part(6, 3)
[[4, 1, 1], [3, 2, 1], [2, 2, 2]]
<强>更新强>
使用memoization:
def part(n, k):
def memoize(f):
cache = [[[None] * n for j in xrange(k)] for i in xrange(n)]
def wrapper(n, k, pre):
if cache[n-1][k-1][pre-1] is None:
cache[n-1][k-1][pre-1] = f(n, k, pre)
return cache[n-1][k-1][pre-1]
return wrapper
@memoize
def _part(n, k, pre):
if n <= 0:
return []
if k == 1:
if n <= pre:
return [(n,)]
return []
ret = []
for i in xrange(min(pre, n), 0, -1):
ret += [(i,) + sub for sub in _part(n-i, k-1, i)]
return ret
return _part(n, k, n)
答案 2 :(得分:2)
首先,我要感谢大家的贡献。 我到达这里需要一个生成整数分区的算法,其中包含以下细节:
将数字的分区生成为完全k个部分,但也具有MINIMUM和MAXIMUM约束。
因此,我修改了&#34; Snakes and Coffee&#34;的代码。满足这些新要求:
def partition_min_max(n,k,l, m):
'''n is the integer to partition, k is the length of partitions,
l is the min partition element size, m is the max partition element size '''
if k < 1:
raise StopIteration
if k == 1:
if n <= m and n>=l :
yield (n,)
raise StopIteration
for i in range(l,m+1):
for result in partition_min_max(n-i,k-1,i,m):
yield result+(i,)
>>> x = list(partition_min_max(20 ,3, 3, 10 ))
>>> print(x)
>>> [(10, 7, 3), (9, 8, 3), (10, 6, 4), (9, 7, 4), (8, 8, 4), (10, 5, 5), (9, 6, 5), (8, 7, 5), (8, 6, 6), (7, 7, 6)]
答案 3 :(得分:0)
基于先前具有最大和最小约束的答案,我们可以将其优化得更好一些。例如,对于 k = 16 , n = 2048 和 m = 128 ,只有一个这样的分区满足约束(128+128+...+128)。但是代码会搜索不必要的分支以寻找可以修剪的答案。
def partition_min_max(n,k,l,m):
#n is the integer to partition, k is the length of partitions,
#l is the min partition element size, m is the max partition element size
if k < 1:
return
if k == 1:
if n <= m and n>=l :
yield (n,)
return
if (k*128) < n: #If the current sum is too small to reach n
return
if k*1 > n:#If current sum is too big to reach n
return
for i in range(l,m+1):
for result in partition_min_max(n-i,k-1,i,m):
yield result+(i,)