题目链接
https://leetcode-cn.com/problems/combination-sum-ii/
解答过程
这个题目和Problem39很像,区别在于两点: (1)Problem39的入参数组无重复元素,Problem40的入参数组可以包含重复元素; (2)Problem39中每个元素可以使用任意次,Problem40中每个元素最多只能使用1次; 稍微思考一下,就能体会出为什么有这样的设定。因为Problem39中每个元素可以使用任意次,所以即便是允许入参数组包含重复元素,那么也得先排序并过滤掉重复元素,题目里限定无重复元素等于是do you a favor. Problem40允许有重复元素但每个元素至多使用1次,其实就可以理解为Problem39的变种,即可以理解为在Problem39的基础上限定了每个元素最多使用几次。
从代码来看,Problem40比Problem39麻烦一些,但是思路是一致的,细节不再赘述。
public List<List<Integer>> combinationSum(int[] candidates, int target) {
Arrays.sort(candidates);
List<int[]> candidatesAndFreq = new ArrayList<>();
Arrays.stream(candidates).forEach(e -> {
if (candidatesAndFreq.size() > 0
&& candidatesAndFreq.get(candidatesAndFreq.size()-1)[0] == e) {
candidatesAndFreq.get(candidatesAndFreq.size() - 1)[1] += 1;
} else {
int[] last = new int[2];
last[0] = e;
last[1] = 1;
candidatesAndFreq.add(last);
}
});
List<List<Integer>> result = new ArrayList<>();
List<Integer> combination = new ArrayList<>();
backtrack(result, combination, candidatesAndFreq, 0, target);
return result;
}
private void backtrack(List<List<Integer>> result, List<Integer> combination,
List<int[]> candidates, int curIdx, int target) {
if (target == 0) {
result.add(new ArrayList<>(combination));
return;
}
if (curIdx == candidates.size()) {
return;
}
int curVal = candidates.get(curIdx)[0];
int freq = candidates.get(curIdx)[1];
if (curVal > target) {
return;
}
for (int i = 0; i <= freq; i++) {
if (target >= i * curVal) {
for (int j = 0; j < i; j++) {
combination.add(curVal);
}
backtrack(result, combination, candidates, curIdx+1, target-i*curVal);
for (int j = 0; j < i; j++) {
combination.remove(combination.size()-1);
}
}
}
}