Problem

Given an integer array arr of size n, return the average (arithmetic mean) of its elements. Return the result as a floating-point number rounded to two decimal places, or as an integer floor if the problem demands integer output. Handle the empty array by returning 0.

Input / Output

• Input: integer array arr.
• Output: the mean.

Examples

arr	Output
[1, 2, 3, 4, 5]	3.0
[10, 20]	15.0
[7]	7.0
[-3, 3]	0.0
[]	0

Constraints

• 0 <= n <= 10^6
• -10^9 <= arr[i] <= 10^9

Approach

Sum all elements and divide by n. The interesting parts are the edge cases the interviewer is really asking about:

• Overflow. With n = 10^6 and values up to 10^9, the sum reaches 10^15, which overflows a 32-bit integer. Use 64-bit (long) or accumulate in floating point.
• Integer vs. floating division. In C, C++ and Java, sum / n with both operands integers truncates. Cast one operand to double first.
• Empty input. Division by zero — return 0 or raise, but decide explicitly.
• Streaming / running average. If values arrive one at a time, keep count and mean and update with mean += (x - mean) / count — this avoids storing the sum and is numerically stabler for very long streams (Welford's method).
• Sliding-window average (average of the last k elements) is the natural follow-up and uses a circular buffer plus a running sum.

Complexity

• Time: O(n)
• Space: O(1)

Reference solution (Python)

def average(arr):
    if not arr:
        return 0
    return round(sum(arr) / len(arr), 2)

def running_mean(stream):
    mean, count = 0.0, 0
    for x in stream:
        count += 1
        mean += (x - mean) / count
    return mean

Test cases

Input	Expected
[1,2,3,4,5]	3.0
[10,20]	15.0
[]	0
running_mean([2,4,6])	4.0