macro_rules! get {
    ($t:ty) => {
        {
            let mut line: String = String::new();
            std::io::stdin().read_line(&mut line).unwrap();
            line.trim().parse::<$t>().unwrap()
        }
    };
    ($($t:ty),*) => {
        {
            let mut line: String = String::new();
            std::io::stdin().read_line(&mut line).unwrap();
            let mut iter = line.split_whitespace();
            (
                $(iter.next().unwrap().parse::<$t>().unwrap(),)*
            )
        }
    };
    ($t:ty; $n:expr) => {
        (0..$n).map(|_|
            get!($t)
        ).collect::<Vec<_>>()
    };
    ($($t:ty),*; $n:expr) => {
        (0..$n).map(|_|
            get!($($t),*)
        ).collect::<Vec<_>>()
    };
    ($t:ty ;;) => {
        {
            let mut line: String = String::new();
            std::io::stdin().read_line(&mut line).unwrap();
            line.split_whitespace()
                .map(|t| t.parse::<$t>().unwrap())
                .collect::<Vec<_>>()
        }
    };
    ($t:ty ;; $n:expr) => {
        (0..$n).map(|_| get!($t ;;)).collect::<Vec<_>>()
    };
}

fn q_sort(nums: Vec<i32>) -> Vec<i32> {
    if nums.len() <= 1 {
        nums
    } else {
        let pivot = nums[0];
        let mut left: Vec<i32> = Vec::new();
        let mut right: Vec<i32> = Vec::new();
        for num in (nums[1..]).into_iter() {
            if *num <= pivot {
                left.push(*num)
            } else {
                right.push(*num)
            }
        }
        let mut sorted_l = q_sort(left);
        let mut sorted_r = q_sort(right);
        let mut res = Vec::new();
        res.append(&mut sorted_l);
        res.push(pivot);
        res.append(&mut sorted_r);
        res
    }
}

fn get_median(nums: Vec<i32>) -> f32 {
    let len = nums.len();
    if len % 2 == 0 {
        (nums[len / 2 - 1] + nums[len / 2]) as f32 / 2.0
    } else {
        nums[len / 2] as f32
    }
}

fn main() {
    let _len = get!(i32);
    let data: Vec<i32> = get!(i32;;);
    println!("{}", get_median(q_sort(data)))
}