Given a list of floating point numbers, print their sum, rounded to the nearest integer.

Input:
Each line will contain a int n (0 < n < 2000), the number of floats on that line, followed by n floating point numbers (0 < f < 2^37). The float will not be more precise than 10^-5. Input will terminate with a line containing a single '0'.

Output:
Output one line per test case, containing the sum of the floats, rounded to the nearest integer.

Input:
4 1 2 3 4.5
2 1.1 8
0

Output:
11
9