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

我在做类似的scala作业。我想在这里发布我的解决方案:

 def countChange(money: Int, coins: List[Int]): Int = {
      def getCount(money: Int, remainingCoins: List[Int]): Int = {
        if(money == 0 ) 1
        else if(money < 0 || remainingCoins.isEmpty) 0
        else
          getCount(money, remainingCoins.tail) +
            getCount(money - remainingCoins.head, remainingCoins)
      }
      if(money == 0 || coins.isEmpty) 0
      else getCount(money, coins)
    }

其他回答

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;
    }
}

PHP版本,灵感来自Keith Beller的c#版本。

bala的PHP版本不适合我,因为我不需要对数字进行分组。我想要一个更简单的实现,只有一个目标值和一个数字池。这个函数也会删除任何重复的条目。

编辑25/10/2021:添加精度参数以支持浮点数(现在需要bcmath扩展)。

/**
 * Calculates a subset sum: finds out which combinations of numbers
 * from the numbers array can be added together to come to the target
 * number.
 *
 * Returns an indexed array with arrays of number combinations.
 *
 * Example:
 *
 * <pre>
 * $matches = subset_sum(array(5,10,7,3,20), 25);
 * </pre>
 *
 * Returns:
 *
 * <pre>
 * Array
 * (
 *   [0] => Array
 *   (
 *       [0] => 3
 *       [1] => 5
 *       [2] => 7
 *       [3] => 10
 *   )
 *   [1] => Array
 *   (
 *       [0] => 5
 *       [1] => 20
 *   )
 * )
 * </pre>
 *
 * @param number[] $numbers
 * @param number $target
 * @param array $part
 * @param int $precision
 * @return array[number[]]
 */
function subset_sum($numbers, $target, $precision=0, $part=null)
{
    // we assume that an empty $part variable means this
    // is the top level call.
    $toplevel = false;
    if($part === null) {
        $toplevel = true;
        $part = array();
    }

    $s = 0;
    foreach($part as $x)
    {
        $s = $s + $x;
    }

    // we have found a match!
    if(bccomp((string) $s, (string) $target, $precision) === 0)
    {
        sort($part); // ensure the numbers are always sorted
        return array(implode('|', $part));
    }

    // gone too far, break off
    if($s >= $target)
    {
        return null;
    }

    $matches = array();
    $totalNumbers = count($numbers);

    for($i=0; $i < $totalNumbers; $i++)
    {
        $remaining = array();
        $n = $numbers[$i];

        for($j = $i+1; $j < $totalNumbers; $j++)
        {
            $remaining[] = $numbers[$j];
        }

        $part_rec = $part;
        $part_rec[] = $n;

        $result = subset_sum($remaining, $target, $precision, $part_rec);
        if($result)
        {
            $matches = array_merge($matches, $result);
        }
    }

    if(!$toplevel)
    {
        return $matches;
    }

    // this is the top level function call: we have to
    // prepare the final result value by stripping any
    // duplicate results.
    $matches = array_unique($matches);
    $result = array();
    foreach($matches as $entry)
    {
        $result[] = explode('|', $entry);
    }

    return $result;
}

例子:

$result = subset_sum(array(5, 10, 7, 3, 20), 25);

这将返回一个包含两个数字组合数组的索引数组:

3, 5, 7, 10
5, 20

浮点数示例:

// Specify the precision in the third argument
$result = subset_sum(array(0.40, 0.03, 0.05), 0.45, 2);

这将返回一个匹配项:

0.40, 0.05

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?

Java解决方案的Swift 3转换(by @JeremyThompson)

protocol _IntType { }
extension Int: _IntType {}


extension Array where Element: _IntType {

    func subsets(to: Int) -> [[Element]]? {

        func sum_up_recursive(_ numbers: [Element], _ target: Int, _ partial: [Element], _ solution: inout [[Element]]) {

            var sum: Int = 0
            for x in partial {
                sum += x as! Int
            }

            if sum == target {
                solution.append(partial)
            }

            guard sum < target else {
                return
            }

            for i in stride(from: 0, to: numbers.count, by: 1) {

                var remaining = [Element]()

                for j in stride(from: i + 1, to: numbers.count, by: 1) {
                    remaining.append(numbers[j])
                }

                var partial_rec = [Element](partial)
                partial_rec.append(numbers[i])

                sum_up_recursive(remaining, target, partial_rec, &solution)
            }
        }

        var solutions = [[Element]]()
        sum_up_recursive(self, to, [Element](), &solutions)

        return solutions.count > 0 ? solutions : nil
    }

}

用法:

let numbers = [3, 9, 8, 4, 5, 7, 10]

if let solution = numbers.subsets(to: 15) {
    print(solution) // output: [[3, 8, 4], [3, 5, 7], [8, 7], [5, 10]]
} else {
    print("not possible")
}

我不喜欢上面看到的Javascript解决方案。下面是我使用部分应用、闭包和递归构建的一个:

好的,我主要关心的是,如果组合数组能满足目标要求,希望这样你就能找到剩下的组合了

这里只需要设置目标并传递组合数组。

function main() {
    const target = 10
    const getPermutationThatSumT = setTarget(target)
    const permutation = getPermutationThatSumT([1, 4, 2, 5, 6, 7])

    console.log( permutation );
}

我提出的当前实现

function setTarget(target) {
    let partial = [];

    return function permute(input) {
        let i, removed;
        for (i = 0; i < input.length; i++) {
            removed = input.splice(i, 1)[0];
            partial.push(removed);

            const sum = partial.reduce((a, b) => a + b)
            if (sum === target) return partial.slice()
            if (sum < target) permute(input)

            input.splice(i, 0, removed);
            partial.pop();
        }
        return null
    };
}