* Seeing this, everyone will definitely think of sorting`nums1` from small to large, so that the front is less than or equal to the back, but the indexes will be**messy** when sorting. If there is no good way to solve this problem, the whole question cannot be solved. Please think about it first.

dddBring the`index` when sorting, that is, the object to be sorted is an array of tuples of`(num, index)`. This technique**must be mastered**, as it will be used in many questions.ddd

<details><summary>Click to view the answer</summary><p>Bring the `index` when sorting, that is, the object to be sorted is an array of tuples of `(num, index)`. This technique **must be mastered**, as it will be used in many questions.</p></details>

After solving the above problems, the indexes are all there, let's continue reading:

* Requirement 2: Choose at most`k` values of`nums2[j]` at these indices to**maximize** the total sum.

After seeing this, have you thought of any good method?

dddHeap sort, maintain a large root heap of size`k`. This is also a knowledge point that is often tested,**must be mastered**.ddd

<details><summary>Click to view the answer</summary><p>Heap sort, maintain a large root heap of size `k`. This is also a knowledge point that is often tested, **must be mastered**.</p></details>

Seeing this, please implement the code according to the above prompts.

* Finally, it is found that the repeated`num` that appear continuously should be specially processed, that is, the values ​​in`answer` corresponding to the same`num` should be the same. There are many ways to deal with it. What is the simplest way to deal with it?

dddUse a`Map`,`key` is`num`, and the same`key` directly uses the`value` corresponding to`key`.ddd

<details><summary>Click to view the answer</summary><p>Use a `Map`, `key` is `num`, and the same `key` directly uses the `value` corresponding to `key`.</p></details>

- 要求1：找出所有满足`nums1[j]` 小于`nums1[i]` 的下标`j` 。

看到这个，大家一定会想到把`nums1` 按从小到大排序，这样，前面的小于等于后面的，但一排序下标就**乱**了。如果对这个问题没有好办法解决，整个题目就做不出来。先请思考下。

ddd在排序时带上索引下标，即排序的对象是元组`(num, index)`的数组。这个技术**一定要掌握**，许多题目都会用到。ddd

<details><summary>点击查看答案</summary><p>在排序时带上索引下标，即排序的对象是元组`(num, index)`的数组。这个技术**一定要掌握**，许多题目都会用到。</p></details>

解决了上面的问题，下标就都有了，我们再看：

- 要求2：从这些下标对应的`nums2[j]` 中选出 至多`k` 个，并**最大化** 这些值的总和作为结果。

看到这个，你想到用什么好方法了吗？

ddd堆排序，维护一个大小为`k` 的大根堆。这也是经常考察的知识点，**一定要掌握**哦。ddd

<details><summary>点击查看答案</summary><p>堆排序，维护一个大小为 `k` 的大根堆。这也是经常考察的知识点，**一定要掌握**哦。</p></details>

看到这，请你先按上文提示把代码实现一下。

- 最后，发现还要对连续出现的重复的`num` 进行特别处理，即相同的`num` 对应的`answer` 中的值应该是一样的。处理方法有多种，怎么处理最简单呢？

ddd用一个`Map`，`key`为`num`, 相同的`key` 直接使用`key` 对应的`值`。ddd

<details><summary>点击查看答案</summary><p>用一个 `Map`， `key`为 `num`, 相同的 `key` 直接使用 `key` 对应的`值`。</p></details>