The first subtask rewards any brute force too slow for the second subtask.

The second subtask rewards brute force solutions.

Time Complexity: ~\mathcal{O}(Th^2)~

The brute force can be optimized with a ~2~-pointer approach for the third subtask.

Time Complexity: ~\mathcal{O}(Th)~

For the final subtask, we can try finding an approximate answer for cases with large ~h~. Observe that this question is similar to counting the number of lattice points under a quarter circle with radius ~h~. Thus, our approximation is ~\frac{\pi h^2}{4}~.

For smaller ~h~, we can run the subtask 3 algorithm so our relative error stays small, and use our approximation for large ~h~.

Time Complexity: ~\mathcal{O}(Th)~