我正在尝试使用递归来编写方法subsetWithSum(ArrayList numbers,int sum),它接受一个整数和整数和的数组,并返回一个包含给定数字(提供的ArrayList)总和为sum。没有必要返回多个组合,如果没有这样的子集,它应该返回null。但是我的代码只为每一个返回null .`
以下是该方法的代码:
public static ArrayList<Integer> subsetWithSum(ArrayList<Integer> numbers, int sum){
ArrayList<Integer> sumList=new ArrayList<Integer>();
int sumForNumbers=0;
for (int i=0; i<=numbers.size()-1; i++)
sumForNumbers+=numbers.get(i);
if (sumForNumbers==sum)
return numbers;
else if(sumForNumbers>sum || numbers.size()==0)
return null;
else {
for (int i=0; i<numbers.size();i++){
int n=numbers.get(i);
for (int currentIndex=i+1; currentIndex<numbers.size(); currentIndex++)
sumList.add(numbers.get(currentIndex));
for (int currentIndex=0; currentIndex<=numbers.size()-1;currentIndex++){
if ((sumForNumbers+numbers.get(currentIndex))<=sum){
sumList.add(numbers.get(currentIndex));
sumForNumbers+=numbers.get(currentIndex);
}
}
}
return subsetWithSum(sumList, sum);
}
}
这是我对main中方法的调用:
public static void main(String[] args) {
ArrayList<Integer> test = new ArrayList<Integer>();
test.add(3); test.add(11); test.add(1); test.add(5);
System.out.println("Available numbers: " +test);
for(int sum=16; sum<=19; sum++){
ArrayList<Integer> answer = subsetWithSum(test, sum);
System.out.println(sum+" can be made with: "+answer);
这是我目前的输出:
Available numbers: [3, 11, 1, 5]`
16 can be made with: null
17 can be made with: null
18 can be made with: null
19 can be made with: null
我的预期输出是:
Available numbers: [3, 11, 1, 5]
16 can be made with: [11, 5]
17 can be made with: [11, 1, 5]
18 can be made with: null
19 can be made with: [3, 11, 5]
我发现递归真的很难理解,任何帮助都会很棒
答案 0 :(得分:0)
首先,如果您使用的是Java 8,那么List<Integer> list
的求和就像list.stream().mapToInt(n -> n).sum()
一样简单。
其次,递归总是采用类似的形式:
func(context)
if context in simple form
return simple result
else
break context down into smaller pieces
call func on smaller pieces
在您的情况下,它看起来像
func(total, list)
if sum(list) == total
return list
else if list is not empty
get all solutions from func(total - first item, list without first item)
and func(total, list without first item)
这里有一些棘手的事情需要考虑:
这是一个带有测试用例的示例解决方案。
public class ListSum {
public static void main(String[] args) {
subsetsThatSumTo(18, Arrays.asList(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)).forEach(System.out::println);
}
public static List<List<Integer>> subsetsThatSumTo(int total, List<Integer> list) {
List<List<Integer>> result = new ArrayList<>();
if (list.stream().mapToInt(n -> n).sum() == total) {
result.add(new ArrayList<>(list));
} else if (!list.isEmpty()) {
subsetsThatSumTo(total - list.get(0), list.subList(1, list.size())).forEach(result::add);
result.forEach(l -> l.add(0, list.get(0)));
subsetsThatSumTo(total, list.subList(1, list.size())).forEach(result::add);
}
return result;
}
}
如果您只想返回第一个结果:
public class ListSum {
public static void main(String[] args) {
System.out.println(subsetThatSumTo(18, Arrays.asList(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)));
}
public static List<Integer> subsetThatSumTo(int total, List<Integer> list) {
if (list.stream().mapToInt(n -> n).sum() == total)
return new ArrayList<>(list);
if (list.isEmpty())
return null;
List<Integer> result = subsetThatSumTo(total - list.get(0), list.subList(1, list.size()));
if (result != null) {
result.add(0, list.get(0));
return result;
} else {
return subsetThatSumTo(total, list.subList(1, list.size()));
}
}
}
答案 1 :(得分:0)
您应该注意的第一件事是这个问题是NP-Complete。这意味着它的运行时间不是多项式的。最着名的算法是指数的。如果您找到一个多项式算法来解决它,并且您可以证明它是正确的,那么您赢得1M $;)
参考: https://en.m.wikipedia.org/wiki/Subset_sum_problem
现在,在这种情况下,您可以应用回溯模式:
https://en.m.wikipedia.org/wiki/Backtracking
只需将伪代码转换为Java并实现这些功能。如果你研究这种模式并练习将它用于不同的问题,你将学习很多关于递归的知识。