Average Pooling 2D

Apply 2-D average pooling with a k×k window and stride s (default 1). Windows that do not fit entirely inside the input are not used. Pass s=k for the usual non-overlapping pooling.

Implement solve(X,k,s=1).

Returns. Return a float array of shape ((H - k)//s + 1, (W - k)//s + 1); if the window is larger than the input the result is empty.

Examples

Example 1

Input:

solve([[1, 3], [2, 4]], 2)

Output:

[[2.5]]

Example 2

Input:

solve([[1, 2, 5, 6], [3, 4, 7, 8], [9, 1, 2, 3], [4, 5, 6, 7]], 2, 2)

Output:

[[2.5, 6.5], [4.75, 4.5]]