Rolling Mean

Compute the trailing rolling mean of width window: position $ii$ holds the mean of the last window values ending at $ii$. The first window - 1 positions do not have a full window yet.

Implement solve(x, window).

Returns. Return a float NumPy array the same length as x, with NaN in the first window - 1 positions. window must be at least 1 (otherwise ValueError); a window longer than x gives all NaN.

Examples

Example 1

Input:

solve([1.0, 2.0, 3.0, 4.0], 2)

Output:

[nan, 1.5, 2.5, 3.5]

Example 2

Input:

solve([2.0, 4.0, 6.0, 8.0, 10.0], 3)

Output:

[nan, nan, 4.0, 6.0, 8.0]