你如何从给定的N个数字中测试所有可能的加法组合，使它们加起来得到给定的最终数字?
一个简单的例子:
要添加的数字集:N ={1,5,22,15,0，…} 期望结果:12345
当前回答
到目前为止，有很多解决方案，但都是生成然后过滤的形式。这意味着他们可能会在递归路径上花费大量时间，而这些递归路径不会导致解决方案。
这里的解决方案是O(size_of_array * (number_of_sum + number_of_solutions))。换句话说，它使用动态规划来避免列举永远不会匹配的可能解决方案。
为了搞笑，我让这个函数同时使用正数和负数，并让它成为一个迭代器。它适用于Python 2.3+。
def subset_sum_iter(array, target):
sign = 1
array = sorted(array)
if target < 0:
array = reversed(array)
sign = -1
# Checkpoint A
last_index = {0: [-1]}
for i in range(len(array)):
for s in list(last_index.keys()):
new_s = s + array[i]
if 0 < (new_s - target) * sign:
pass # Cannot lead to target
elif new_s in last_index:
last_index[new_s].append(i)
else:
last_index[new_s] = [i]
# Checkpoint B
# Now yield up the answers.
def recur(new_target, max_i):
for i in last_index[new_target]:
if i == -1:
yield [] # Empty sum.
elif max_i <= i:
break # Not our solution.
else:
for answer in recur(new_target - array[i], i):
answer.append(array[i])
yield answer
for answer in recur(target, len(array)):
yield answer
这里有一个例子，它与数组和目标一起使用，在其他解决方案中使用的过滤方法实际上永远不会结束。
def is_prime(n):
for i in range(2, n):
if 0 == n % i:
return False
elif n < i * i:
return True
if n == 2:
return True
else:
return False
def primes(limit):
n = 2
while True:
if is_prime(n):
yield(n)
n = n + 1
if limit < n:
break
for answer in subset_sum_iter(primes(1000), 76000):
print(answer)
这将在2秒内打印所有522个答案。之前的方法如果能在宇宙当前的生命周期内找到答案，那就太幸运了。(整个空间有2^168 = 3.74144419156711e+50个可能的组合。那需要一段时间。)
解释 我被要求解释代码，但解释数据结构通常更能说明问题。我来解释一下数据结构。
让我们考虑subset_sum_iter([2, 2、3、3、5、5、7、7、-11、11),10)。
在检查点A，我们已经意识到我们的目标是正的，所以符号= 1。我们已经对输入进行了排序，使array =[-11， -7， -5， -3， -2, 2,3,5,7,11]。由于我们经常通过索引访问它，下面是从索引到值的映射:
0: -11
1: -7
2: -5
3: -3
4: -2
5: 2
6: 3
7: 5
8: 7
9: 11
通过检查点B，我们使用动态规划生成last_index数据结构。它包含什么?
last_index = {
-28: [4],
-26: [3, 5],
-25: [4, 6],
-24: [5],
-23: [2, 4, 5, 6, 7],
-22: [6],
-21: [3, 4, 5, 6, 7, 8],
-20: [4, 6, 7],
-19: [3, 5, 7, 8],
-18: [1, 4, 5, 6, 7, 8],
-17: [4, 5, 6, 7, 8, 9],
-16: [2, 4, 5, 6, 7, 8],
-15: [3, 5, 6, 7, 8, 9],
-14: [3, 4, 5, 6, 7, 8, 9],
-13: [4, 5, 6, 7, 8, 9],
-12: [2, 4, 5, 6, 7, 8, 9],
-11: [0, 5, 6, 7, 8, 9],
-10: [3, 4, 5, 6, 7, 8, 9],
-9: [4, 5, 6, 7, 8, 9],
-8: [3, 5, 6, 7, 8, 9],
-7: [1, 4, 5, 6, 7, 8, 9],
-6: [5, 6, 7, 8, 9],
-5: [2, 4, 5, 6, 7, 8, 9],
-4: [6, 7, 8, 9],
-3: [3, 5, 6, 7, 8, 9],
-2: [4, 6, 7, 8, 9],
-1: [5, 7, 8, 9],
0: [-1, 5, 6, 7, 8, 9],
1: [6, 7, 8, 9],
2: [5, 6, 7, 8, 9],
3: [6, 7, 8, 9],
4: [7, 8, 9],
5: [6, 7, 8, 9],
6: [7, 8, 9],
7: [7, 8, 9],
8: [7, 8, 9],
9: [8, 9],
10: [7, 8, 9]
}
(旁注，它不是对称的，因为条件if 0 < (new_s - target) *符号阻止我们记录超过target的任何内容，在我们的例子中是10。)
这是什么意思?以条目10为例:[7,8,9]。这意味着我们可以得到10的最终和，最后选择的数字在索引7、8或9处。也就是说，最后选择的数字可以是5,7或11。
让我们仔细看看如果我们选择索引7会发生什么。这意味着我们以5结束。因此，在得到下标7之前，我们必须得到10-5 = 5。5的条目为5:[6,7,8,9]。所以我们可以选择指数6，也就是3。虽然我们在第7、8和9处得到了5，但在第7号下标之前我们没有得到5。所以倒数第二个选项是指数6处的3。
现在我们要在下标6之前得到5-3 = 2。条目2是:2:[5,6,7,8,9]。同样，我们只关心下标5的答案因为其他的都发生得太晚了。所以倒数第三个选项是指数5处的2。
最后我们要在下标5之前得到2-2 = 0。条目0表示:0:[- 1,5,6,7,8,9]。同样，我们只关心-1。但是-1不是下标实际上我用它来表示我们已经完成了选择。
我们求出了2+3+5 = 10的解。这是我们打印出来的第一个解。
现在我们来看递归子函数。因为它是在main函数内部定义的，所以它可以看到last_index。
首先要注意的是，它调用的是yield，而不是return。这使它成为一个发电机。当你调用它时，你会返回一个特殊类型的迭代器。当你循环遍历那个迭代器时，你会得到一个它能产生的所有东西的列表。但你是在生成它们时得到它们的。如果它是一个很长的列表，你不把它放在内存中。(有点重要，因为我们可以得到一个很长的列表。)
recur(new_target, max_i)将产生的结果是你可以用数组中最大索引为max_i的元素求和为new_target的所有方法。这就是它的答案:“我们必须在索引max_i+1之前到达new_target。”当然，它是递归的。
因此，recur(target, len(array))是所有使用任意索引到达目标的解。这就是我们想要的。
其他回答
在Haskell:
filter ((==) 12345 . sum) $ subsequences [1,5,22,15,0,..]
J:
(]#~12345=+/@>)(]<@#~[:#:@i.2^#)1 5 22 15 0 ...
正如您可能注意到的，两者都采用相同的方法，并将问题分为两部分:生成幂集的每个成员，并检查每个成员与目标的和。
还有其他的解决方案，但这是最直接的。
在这两种方法中，你是否需要帮助，或者找到另一种方法?
下面是一个Java版本，它非常适合小N和非常大的目标和，当复杂度O(t*N)(动态解)大于指数算法时。我的版本在中间攻击中使用了一个meet，并进行了一些调整，以降低复杂度，从经典的naive O(n*2^n)降低到O(2^(n/2))。
如果你想在32到64个元素之间的集合中使用这种方法，你应该将表示step函数中当前子集的int改为long，尽管随着集合大小的增加，性能显然会急剧下降。如果你想对一个有奇数个元素的集合使用这个，你应该给这个集合加上一个0，使它成为偶数。
import java.util.ArrayList;
import java.util.List;
public class SubsetSumMiddleAttack {
static final int target = 100000000;
static final int[] set = new int[]{ ... };
static List<Subset> evens = new ArrayList<>();
static List<Subset> odds = new ArrayList<>();
static int[][] split(int[] superSet) {
int[][] ret = new int[2][superSet.length / 2];
for (int i = 0; i < superSet.length; i++) ret[i % 2][i / 2] = superSet[i];
return ret;
}
static void step(int[] superSet, List<Subset> accumulator, int subset, int sum, int counter) {
accumulator.add(new Subset(subset, sum));
if (counter != superSet.length) {
step(superSet, accumulator, subset + (1 << counter), sum + superSet[counter], counter + 1);
step(superSet, accumulator, subset, sum, counter + 1);
}
}
static void printSubset(Subset e, Subset o) {
String ret = "";
for (int i = 0; i < 32; i++) {
if (i % 2 == 0) {
if ((1 & (e.subset >> (i / 2))) == 1) ret += " + " + set[i];
}
else {
if ((1 & (o.subset >> (i / 2))) == 1) ret += " + " + set[i];
}
}
if (ret.startsWith(" ")) ret = ret.substring(3) + " = " + (e.sum + o.sum);
System.out.println(ret);
}
public static void main(String[] args) {
int[][] superSets = split(set);
step(superSets[0], evens, 0,0,0);
step(superSets[1], odds, 0,0,0);
for (Subset e : evens) {
for (Subset o : odds) {
if (e.sum + o.sum == target) printSubset(e, o);
}
}
}
}
class Subset {
int subset;
int sum;
Subset(int subset, int sum) {
this.subset = subset;
this.sum = sum;
}
}
func sum(array : [Int]) -> Int{
var sum = 0
array.forEach { (item) in
sum = item + sum
}
return sum
}
func susetNumbers(array :[Int], target : Int, subsetArray: [Int],result : inout [[Int]]) -> [[Int]]{
let s = sum(array: subsetArray)
if(s == target){
print("sum\(subsetArray) = \(target)")
result.append(subsetArray)
}
for i in 0..<array.count{
let n = array[i]
let remaning = Array(array[(i+1)..<array.count])
susetNumbers(array: remaning, target: target, subsetArray: subsetArray + [n], result: &result)
}
return result
}
var resultArray = [[Int]]()
let newA = susetNumbers(array: [1,2,3,4,5], target: 5, subsetArray: [],result:&resultArray)
print(resultArray)
Java非递归版本，简单地添加元素并在可能的值之间重新分配它们。0被忽略，适用于固定的列表(给定的是您可以使用的)或可重复的数字列表。
import java.util.*;
public class TestCombinations {
public static void main(String[] args) {
ArrayList<Integer> numbers = new ArrayList<>(Arrays.asList(0, 1, 2, 2, 5, 10, 20));
LinkedHashSet<Integer> targets = new LinkedHashSet<Integer>() {{
add(4);
add(10);
add(25);
}};
System.out.println("## each element can appear as many times as needed");
for (Integer target: targets) {
Combinations combinations = new Combinations(numbers, target, true);
combinations.calculateCombinations();
for (String solution: combinations.getCombinations()) {
System.out.println(solution);
}
}
System.out.println("## each element can appear only once");
for (Integer target: targets) {
Combinations combinations = new Combinations(numbers, target, false);
combinations.calculateCombinations();
for (String solution: combinations.getCombinations()) {
System.out.println(solution);
}
}
}
public static class Combinations {
private boolean allowRepetitions;
private int[] repetitions;
private ArrayList<Integer> numbers;
private Integer target;
private Integer sum;
private boolean hasNext;
private Set<String> combinations;
/**
* Constructor.
*
* @param numbers Numbers that can be used to calculate the sum.
* @param target Target value for sum.
*/
public Combinations(ArrayList<Integer> numbers, Integer target) {
this(numbers, target, true);
}
/**
* Constructor.
*
* @param numbers Numbers that can be used to calculate the sum.
* @param target Target value for sum.
*/
public Combinations(ArrayList<Integer> numbers, Integer target, boolean allowRepetitions) {
this.allowRepetitions = allowRepetitions;
if (this.allowRepetitions) {
Set<Integer> numbersSet = new HashSet<>(numbers);
this.numbers = new ArrayList<>(numbersSet);
} else {
this.numbers = numbers;
}
this.numbers.removeAll(Arrays.asList(0));
Collections.sort(this.numbers);
this.target = target;
this.repetitions = new int[this.numbers.size()];
this.combinations = new LinkedHashSet<>();
this.sum = 0;
if (this.repetitions.length > 0)
this.hasNext = true;
else
this.hasNext = false;
}
/**
* Calculate and return the sum of the current combination.
*
* @return The sum.
*/
private Integer calculateSum() {
this.sum = 0;
for (int i = 0; i < repetitions.length; ++i) {
this.sum += repetitions[i] * numbers.get(i);
}
return this.sum;
}
/**
* Redistribute picks when only one of each number is allowed in the sum.
*/
private void redistribute() {
for (int i = 1; i < this.repetitions.length; ++i) {
if (this.repetitions[i - 1] > 1) {
this.repetitions[i - 1] = 0;
this.repetitions[i] += 1;
}
}
if (this.repetitions[this.repetitions.length - 1] > 1)
this.repetitions[this.repetitions.length - 1] = 0;
}
/**
* Get the sum of the next combination. When 0 is returned, there's no other combinations to check.
*
* @return The sum.
*/
private Integer next() {
if (this.hasNext && this.repetitions.length > 0) {
this.repetitions[0] += 1;
if (!this.allowRepetitions)
this.redistribute();
this.calculateSum();
for (int i = 0; i < this.repetitions.length && this.sum != 0; ++i) {
if (this.sum > this.target) {
this.repetitions[i] = 0;
if (i + 1 < this.repetitions.length) {
this.repetitions[i + 1] += 1;
if (!this.allowRepetitions)
this.redistribute();
}
this.calculateSum();
}
}
if (this.sum.compareTo(0) == 0)
this.hasNext = false;
}
return this.sum;
}
/**
* Calculate all combinations whose sum equals target.
*/
public void calculateCombinations() {
while (this.hasNext) {
if (this.next().compareTo(target) == 0)
this.combinations.add(this.toString());
}
}
/**
* Return all combinations whose sum equals target.
*
* @return Combinations as a set of strings.
*/
public Set<String> getCombinations() {
return this.combinations;
}
@Override
public String toString() {
StringBuilder stringBuilder = new StringBuilder("" + sum + ": ");
for (int i = 0; i < repetitions.length; ++i) {
for (int j = 0; j < repetitions[i]; ++j) {
stringBuilder.append(numbers.get(i) + " ");
}
}
return stringBuilder.toString();
}
}
}
样例输入:
numbers: 0, 1, 2, 2, 5, 10, 20
targets: 4, 10, 25
样例输出:
## each element can appear as many times as needed
4: 1 1 1 1
4: 1 1 2
4: 2 2
10: 1 1 1 1 1 1 1 1 1 1
10: 1 1 1 1 1 1 1 1 2
10: 1 1 1 1 1 1 2 2
10: 1 1 1 1 2 2 2
10: 1 1 2 2 2 2
10: 2 2 2 2 2
10: 1 1 1 1 1 5
10: 1 1 1 2 5
10: 1 2 2 5
10: 5 5
10: 10
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2
25: 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2
25: 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2
25: 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2
25: 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2
25: 1 1 1 2 2 2 2 2 2 2 2 2 2 2
25: 1 2 2 2 2 2 2 2 2 2 2 2 2
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 5
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 5
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 5
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 5
25: 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 5
25: 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 5
25: 1 1 1 1 1 1 1 1 2 2 2 2 2 2 5
25: 1 1 1 1 1 1 2 2 2 2 2 2 2 5
25: 1 1 1 1 2 2 2 2 2 2 2 2 5
25: 1 1 2 2 2 2 2 2 2 2 2 5
25: 2 2 2 2 2 2 2 2 2 2 5
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 5 5
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 2 5 5
25: 1 1 1 1 1 1 1 1 1 1 1 2 2 5 5
25: 1 1 1 1 1 1 1 1 1 2 2 2 5 5
25: 1 1 1 1 1 1 1 2 2 2 2 5 5
25: 1 1 1 1 1 2 2 2 2 2 5 5
25: 1 1 1 2 2 2 2 2 2 5 5
25: 1 2 2 2 2 2 2 2 5 5
25: 1 1 1 1 1 1 1 1 1 1 5 5 5
25: 1 1 1 1 1 1 1 1 2 5 5 5
25: 1 1 1 1 1 1 2 2 5 5 5
25: 1 1 1 1 2 2 2 5 5 5
25: 1 1 2 2 2 2 5 5 5
25: 2 2 2 2 2 5 5 5
25: 1 1 1 1 1 5 5 5 5
25: 1 1 1 2 5 5 5 5
25: 1 2 2 5 5 5 5
25: 5 5 5 5 5
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 10
25: 1 1 1 1 1 1 1 1 1 1 1 1 1 2 10
25: 1 1 1 1 1 1 1 1 1 1 1 2 2 10
25: 1 1 1 1 1 1 1 1 1 2 2 2 10
25: 1 1 1 1 1 1 1 2 2 2 2 10
25: 1 1 1 1 1 2 2 2 2 2 10
25: 1 1 1 2 2 2 2 2 2 10
25: 1 2 2 2 2 2 2 2 10
25: 1 1 1 1 1 1 1 1 1 1 5 10
25: 1 1 1 1 1 1 1 1 2 5 10
25: 1 1 1 1 1 1 2 2 5 10
25: 1 1 1 1 2 2 2 5 10
25: 1 1 2 2 2 2 5 10
25: 2 2 2 2 2 5 10
25: 1 1 1 1 1 5 5 10
25: 1 1 1 2 5 5 10
25: 1 2 2 5 5 10
25: 5 5 5 10
25: 1 1 1 1 1 10 10
25: 1 1 1 2 10 10
25: 1 2 2 10 10
25: 5 10 10
25: 1 1 1 1 1 20
25: 1 1 1 2 20
25: 1 2 2 20
25: 5 20
## each element can appear only once
4: 2 2
10: 1 2 2 5
10: 10
25: 1 2 2 20
25: 5 20
我将c#示例移植到Objective-c，并没有在响应中看到它:
//Usage
NSMutableArray* numberList = [[NSMutableArray alloc] init];
NSMutableArray* partial = [[NSMutableArray alloc] init];
int target = 16;
for( int i = 1; i<target; i++ )
{ [numberList addObject:@(i)]; }
[self findSums:numberList target:target part:partial];
//*******************************************************************
// Finds combinations of numbers that add up to target recursively
//*******************************************************************
-(void)findSums:(NSMutableArray*)numbers target:(int)target part:(NSMutableArray*)partial
{
int s = 0;
for (NSNumber* x in partial)
{ s += [x intValue]; }
if (s == target)
{ NSLog(@"Sum[%@]", partial); }
if (s >= target)
{ return; }
for (int i = 0;i < [numbers count];i++ )
{
int n = [numbers[i] intValue];
NSMutableArray* remaining = [[NSMutableArray alloc] init];
for (int j = i + 1; j < [numbers count];j++)
{ [remaining addObject:@([numbers[j] intValue])]; }
NSMutableArray* partRec = [[NSMutableArray alloc] initWithArray:partial];
[partRec addObject:@(n)];
[self findSums:remaining target:target part:partRec];
}
}
