Exercise: values above the column mean

Write a function above_mean_counts(X): it takes a two-dimensional numpy array and returns a one-dimensional array. For each column of X, it holds the number of values in that column that are greater than the column's mean.

• above_mean_counts(np.array([[1, 10], [5, 20], [6, 90]])) → [2, 1]
• above_mean_counts(np.array([[1], [2], [3]])) → [1]
• above_mean_counts(np.array([[4, 1], [4, 2], [4, 6]])) → [0, 1]

Constraints: X has at least one row, and its values may be negative or fractional; a value equal to the mean is not counted; the result's length equals the number of columns.

The "Run tests" button runs the checks from the test_above_mean.py tab—you can read them but not change them.

Do not add extra imports unless the task says so explicitly.