You are given an integer N. We call a string S good if it satisfies the following conditions:

\(S[i] = a \text{ or } S[i] = b \text{ } \forall \text{ } 1 \leq i \leq len(S)\)
\(S[i] \leq S[i + 1] \text{ } \forall \text{ } 1\leq i \leq len(S) - 1\). Recall that \(a < b\) according to the numbering of ASCII table.

Here \(len(S)\) denotes the length of string $$S$$. For example, the strings "\(aabb\)", "\(abb\)", "\(aaa\)", "\(bbb\)", "\(a\)" are good but strings "\(aba\)", "\(bab\)", "\(bba\)" are not.

Your task is to find the minimum length of a good string which has atleast N distinct substrings.

Notes

A substring of a string is a contiguous subsequence of that string. For example, "hacker" is subtring of "hackerearth", but "hackr" is not.
Two substrings \(X\) and \(Y\) are different if:
- They are of different length.
- If they are of the same length, then there is an index \(i \) such that \(X[i] \neq Y[i]\).

Input

The first line contains a single integer T denoting the number of test cases.
For each test case, the only line of input contains a single integer N.

Output

For each test case (in a separate line), output the minimum length of a good string which has at least N distinct substrings.

Constraints

\(1 ≤ T ≤ 5000 \\ 1 ≤ N ≤ 10^9\)